SQL Server Big Data Clusters : Data

Virtualization, Data Lake, and AI Platform

Benjamin Weissman download

https://textbookfull.com/product/sql-server-big-data-clusters-
data-virtualization-data-lake-and-ai-platform-benjamin-weissman/

Download more ebook from https://textbookfull.com

We believe these products will be a great fit for you. Click
the link to download now, or visit textbookfull.com
to discover even more!

SQL Server 2019 Revealed: Including Big Data Clusters

and Machine Learning Bob Ward

https://textbookfull.com/product/sql-server-2019-revealed-
including-big-data-clusters-and-machine-learning-bob-ward/

Biota Grow 2C gather 2C cook Loucas

https://textbookfull.com/product/biota-grow-2c-gather-2c-cook-
loucas/

Quick Start Guide to Azure Data Factory Azure Data Lake

Server and Azure Data Warehouse 1st Edition Mark
Beckner

https://textbookfull.com/product/quick-start-guide-to-azure-data-
factory-azure-data-lake-server-and-azure-data-warehouse-1st-
edition-mark-beckner/

SQL Server Data Automation Through Frameworks: Building

Metadata-Driven Frameworks with T-SQL, SSIS, and Azure
Data Factory Andy Leonard

https://textbookfull.com/product/sql-server-data-automation-
through-frameworks-building-metadata-driven-frameworks-with-t-
sql-ssis-and-azure-data-factory-andy-leonard/
Data Lake Analytics on Microsoft Azure: A
Practitioner's Guide to Big Data Engineering Harsh
Chawla

https://textbookfull.com/product/data-lake-analytics-on-
microsoft-azure-a-practitioners-guide-to-big-data-engineering-
harsh-chawla/

Data Analytics for Business AI-ML-PBI-SQL-R Wolfgang

Garn

https://textbookfull.com/product/data-analytics-for-business-ai-
ml-pbi-sql-r-wolfgang-garn/

Data Mining and Big Data Ying Tan

https://textbookfull.com/product/data-mining-and-big-data-ying-
tan/

Health 4 0 How Virtualization and Big Data are

Revolutionizing Healthcare 1st Edition Christoph
Thuemmler

https://textbookfull.com/product/health-4-0-how-virtualization-
and-big-data-are-revolutionizing-healthcare-1st-edition-
christoph-thuemmler/

Data Lake Architecture Designing the Data Lake and

Avoiding the Garbage Dump First Edition Bill Inmon

https://textbookfull.com/product/data-lake-architecture-
designing-the-data-lake-and-avoiding-the-garbage-dump-first-
edition-bill-inmon/
SQL Server
Big Data Clusters
Data Virtualization, Data Lake, and
AI Platform
—
Second Edition
—
Benjamin Weissman
Enrico van de Laar
 SQL Server Big Data
Clusters
Data Virtualization, Data Lake,
and AI Platform
Second Edition

Benjamin Weissman
Enrico van de Laar
SQL Server Big Data Clusters: Data Virtualization, Data Lake, and AI Platform
Benjamin Weissman Enrico van de Laar
Nurnberg, Germany Drachten, The Netherlands

ISBN-13 (pbk): 978-1-4842-5984-9 ISBN-13 (electronic): 978-1-4842-5985-6

https://doi.org/10.1007/978-1-4842-5985-6

Copyright © 2020 by Benjamin Weissman and Enrico van de Laar

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the
material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now
known or hereafter developed.
Trademarked names, logos, and images may appear in this book. Rather than use a trademark symbol with
every occurrence of a trademarked name, logo, or image we use the names, logos, and images only in an
editorial fashion and to the benefit of the trademark owner, with no intention of infringement of the
trademark.
The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not
identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to
proprietary rights.
While the advice and information in this book are believed to be true and accurate at the date of publication,
neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or
omissions that may be made. The publisher makes no warranty, express or implied, with respect to the
material contained herein.
Managing Director, Apress Media LLC: Welmoed Spahr
Acquisitions Editor: Jonathan Gennick
Development Editor: Laura Berendson
Coordinating Editor: Jill Balzano
Cover image designed by Freepik (www.freepik.com)
Distributed to the book trade worldwide by Springer Science+Business Media New York, 233 Spring Street,
6th Floor, New York, NY 10013. Phone 1-800-SPRINGER, fax (201) 348-4505, e-mail orders-ny@springer-
sbm.com, or visit www.springeronline.com. Apress Media, LLC is a California LLC and the sole member
(owner) is Springer Science + Business Media Finance Inc (SSBM Finance Inc). SSBM Finance Inc is a
Delaware corporation.
For information on translations, please e-mail rights@apress.com, or visit http://www.apress.com/
rights-permissions.
Apress titles may be purchased in bulk for academic, corporate, or promotional use. eBook versions and
licenses are also available for most titles. For more information, reference our Print and eBook Bulk Sales
web page at http://www.apress.com/bulk-sales.
Any source code or other supplementary material referenced by the author in this book is available to
readers on GitHub via the book's product page, located at www.apress.com/9781484259849. For more
detailed information, please visit http://www.apress.com/source-code.
Printed on acid-free paper
This one is dedicated to all the ravers in the nation.
Table of Contents
About the Authors���������������������������������������������������������������������������������������������������� ix

About the Technical Reviewer��������������������������������������������������������������������������������� xi

Acknowledgments������������������������������������������������������������������������������������������������� xiii

Introduction�������������������������������������������������������������������������������������������������������������xv

Chapter 1: What Are Big Data Clusters?������������������������������������������������������������������� 1

What Is a SQL Server 2019 Big Data Cluster Really?�������������������������������������������������������������������� 2
Data Virtualization������������������������������������������������������������������������������������������������������������������������� 2
Outsource Your Data���������������������������������������������������������������������������������������������������������������� 6
Reduce Data Redundancy and Development Time������������������������������������������������������������������ 6
A Combined Data Platform Environment��������������������������������������������������������������������������������������� 7
Centralized AI Platform����������������������������������������������������������������������������������������������������������������� 9

Chapter 2: Big Data Cluster Architecture��������������������������������������������������������������� 11

Physical Big Data Cluster Infrastructure������������������������������������������������������������������������������������� 11
Containers����������������������������������������������������������������������������������������������������������������������������� 11
SQL Server on Linux�������������������������������������������������������������������������������������������������������������� 15
Spark������������������������������������������������������������������������������������������������������������������������������������� 18
HDFS�������������������������������������������������������������������������������������������������������������������������������������� 22
Tying the Physical Infrastructure Parts Together������������������������������������������������������������������� 23
Logical Big Data Cluster Architecture����������������������������������������������������������������������������������������� 25
Control Plane������������������������������������������������������������������������������������������������������������������������� 27
Compute Area������������������������������������������������������������������������������������������������������������������������ 28
Data Area������������������������������������������������������������������������������������������������������������������������������� 28
Summary������������������������������������������������������������������������������������������������������������������������������������ 31

v
Table of Contents

Chapter 3: Deployment of Big Data Clusters���������������������������������������������������������� 33

A Little Helper: Chocolatey���������������������������������������������������������������������������������������������������������� 33
Installation of an On-Premises PolyBase Instance���������������������������������������������������������������������� 35
Using Azure Data Studio to Work with Big Data Clusters������������������������������������������������������������ 51
What Is Azure Data Studio?��������������������������������������������������������������������������������������������������� 52
Getting and Installing Azure Data Studio������������������������������������������������������������������������������� 52
Installation of a “Real” Big Data Cluster������������������������������������������������������������������������������������� 53
kubeadm on Linux����������������������������������������������������������������������������������������������������������������� 53
Azure Kubernetes Service (AKS)������������������������������������������������������������������������������������������� 56
Deploy Your Big Data Cluster Through Azure Data Studio����������������������������������������������������� 63
What Is azdata?��������������������������������������������������������������������������������������������������������������������� 72
Others������������������������������������������������������������������������������������������������������������������������������������ 81
Advanced Deployment Options��������������������������������������������������������������������������������������������������� 81
Active Directory Authentication for Big Data Clusters����������������������������������������������������������� 82
HDFS Tiering in Big Data Clusters����������������������������������������������������������������������������������������� 82
Summary������������������������������������������������������������������������������������������������������������������������������������ 83

Chapter 4: Loading Data into Big Data Clusters����������������������������������������������������� 85

Getting Azure Data Studio Fully Ready for Your Big Data Clusters���������������������������������������������� 85
Getting Some Sample Files into the Installation������������������������������������������������������������������������� 89
Empty Database�������������������������������������������������������������������������������������������������������������������� 89
Sample Data Within Your Big Data Cluster���������������������������������������������������������������������������� 89
Azure SQL Database�������������������������������������������������������������������������������������������������������������� 95
Summary���������������������������������������������������������������������������������������������������������������������������������� 104

Chapter 5: Querying Big Data Clusters Through T-SQL����������������������������������������� 105

External Tables�������������������������������������������������������������������������������������������������������������������������� 105
Automated External Tables with Biml���������������������������������������������������������������������������������� 118
External Tables from CSV Files in HDFS������������������������������������������������������������������������������ 123
Accessing Data in an Azure Blob Storage���������������������������������������������������������������������������� 138
External Tables from Other Data Sources��������������������������������������������������������������������������������� 138
File-Based Data Sources����������������������������������������������������������������������������������������������������� 138
ODBC����������������������������������������������������������������������������������������������������������������������������������� 140
vi
 Table of Contents

Others���������������������������������������������������������������������������������������������������������������������������������� 140
The SqlDataPool������������������������������������������������������������������������������������������������������������������ 141
Indexes on the SqlDataPool������������������������������������������������������������������������������������������������� 143
Summary���������������������������������������������������������������������������������������������������������������������������������� 145

Chapter 6: Working with Spark in Big Data Clusters�������������������������������������������� 147

Loading Data and Creating a Spark Notebook�������������������������������������������������������������������������� 148
Working with Spark Data Frames��������������������������������������������������������������������������������������������� 151
More Advanced Data Frame Handling��������������������������������������������������������������������������������������� 161
Working with SQL Queries on Spark Data Frames�������������������������������������������������������������������� 173
Reading Data from the SQL Server Master Instance���������������������������������������������������������������� 175
Plotting Graphs������������������������������������������������������������������������������������������������������������������������� 178
Data Frame Execution��������������������������������������������������������������������������������������������������������������� 188
Data Frame Caching������������������������������������������������������������������������������������������������������������������ 190
Data Frame Partitioning������������������������������������������������������������������������������������������������������������ 198
Summary���������������������������������������������������������������������������������������������������������������������������������� 202

Chapter 7: Machine Learning on Big Data Clusters���������������������������������������������� 203

SQL Server In-Database Machine Learning Services��������������������������������������������������������������� 204
Training Machine Learning Models in the SQL Server Master Instance������������������������������ 205
Scoring Data Using In-Database Machine Learning Models����������������������������������������������� 213
Machine Learning in Spark������������������������������������������������������������������������������������������������������� 217
Summary���������������������������������������������������������������������������������������������������������������������������������� 224

Chapter 8: Create and Consume Big Data Cluster Apps��������������������������������������� 225

Create a Big Data Cluster App��������������������������������������������������������������������������������������������������� 226
Consume Big Data Cluster Apps Through REST API������������������������������������������������������������������ 233
Summary���������������������������������������������������������������������������������������������������������������������������������� 238

Chapter 9: Maintenance of Big Data Clusters������������������������������������������������������� 239

Checking the Status of a Big Data Cluster�������������������������������������������������������������������������������� 239
Retrieving a Big Data Cluster’s Status Using azdata����������������������������������������������������������� 239
Manage a Big Data Cluster Using ADS�������������������������������������������������������������������������������� 241

vii
Table of Contents

Metrics (Grafana)����������������������������������������������������������������������������������������������������������������� 244

Log Search Analytics (Kibana)��������������������������������������������������������������������������������������������� 246
Troubleshooting Big Data Clusters�������������������������������������������������������������������������������������������� 247
Upgrading Big Data Clusters����������������������������������������������������������������������������������������������������� 249
Removing a Big Data Cluster Instance�������������������������������������������������������������������������������������� 251
Summary���������������������������������������������������������������������������������������������������������������������������������� 252

Index��������������������������������������������������������������������������������������������������������������������� 253

viii
About the Authors

Benjamin Weissman is the owner and founder of Solisyon, a consulting firm based in
Germany and focused on business intelligence, business analytics, data warehousing, as
well as forecasting and budgeting. He is a Microsoft Data Platform MVP, the first German
BimlHero, and has been working with SQL Server since SQL Server 6.5. If he’s not
currently working with data, he is probably traveling and exploring the world, running,
or enjoying delicious food. You can find Ben on Twitter at @bweissman.

Enrico van de Laar has been working with data in various formats and sizes for over
15 years. He is a data and advanced analytics consultant at Dataheroes where he helps
organizations get the most out of their data. Enrico is a Microsoft Data Platform MVP
since 2014 and a frequent speaker at various data-related events all over the world.
He writes about a wide variety of Microsoft data-related technologies on his blog at
enricovandelaar.com. You can reach Enrico on Twitter at @evdlaar.

ix
About the Technical Reviewer
Mohammad Darab is a data professional with over
20 years of IT experience, 10 years of that working with SQL
Server. He’s a speaker, blogger, and a self-proclaimed Big
Data Cluster advocate. Since the introduction of Big Data
Clusters in SQL Server 2019, Mohammad has been actively
advocating what he calls “the future of SQL Server” through
his social media outlets and blog at MohammadDarab.com.
When he’s not creating Big Data Cluster content, he’s
spending time with his wife and their three kids in their
home in Virginia.

xi
Acknowledgments
As with every publication, a big THANK YOU goes to our families for the support they
gave us during this time-consuming process!
Also, thank you very much to Mohammad for his support by reviewing this book!
We would also like to thank the Microsoft SQL Server product team for helping us
out whenever we had a question or ran into situations we didn’t quite understand. JRJ,
Travis, Buck, Mihaela, and all the others – you rock!
Last but not least, thank you #sqlfamily – your ongoing support, feedback, and
motivation is what keeps us going when it comes to exploring and talking about exciting
technologies like Big Data Clusters!

xiii
Introduction
When we first started talking about writing a book about SQL Server Big Data Clusters,
it was still in one of its first iterations. We both were very excited about all the
technologies included in the product and the way it could potentially change the field
of data processing and analytics. Little did we know how much changes the product
was going to receive while we were writing this. Ultimately this resulted in us almost
rewriting the entire book on a monthly basis. While this was a massive endeavor, it also
allowed us to follow, and document, everything the product went through during its
development. Now that the final product has shipped, we thought it was about time to
provide an updated version that reflects everything that Big Data Clusters is today; the
result is the book in front of you right now!
SQL Server Big Data Clusters is an incredibly exciting new platform. As mentioned
earlier, it consists of a wide variety of different technologies that make it work.
Kubernetes, HDFS, Spark, and SQL Server on Linux are just some of the major players
inside a Big Data Cluster. Besides all these different products combined into a single
product, you can also deploy it on-premises or in the Azure cloud depending on your
use case. As you can imagine, it is near impossible for a single book to discuss all these
different products in depth (as a matter of fact, there are plenty of books available
that do go into all the tiny details for each individual product that is part of a Big Data
Cluster like Spark or SQL Server on Linux). For this reason, we have opted for a different
approach for this book and will focus more on the architecture of Big Data Clusters in
general and practical examples on how to leverage the different approaches on data
processing and analytics Big Data Clusters offer.
With this approach, we believe that while you read this book, you will be able to
understand what makes Big Data Clusters tick, what their use cases are, and how to get
started with deploying, managing, and working with a Big Data Cluster. In that manner
this book tries to deliver useful information that can be used for the various job roles
that deal with data – from data architects that would like more information on how Big
Data Clusters can serve as a centralized data hub to database administrators that want
to know how to manage and deploy databases to the cluster, data scientists that want to
train and operationalize machine learning models on the Big Data Cluster, and many
more different roles. If you are working with data in any way, this book should have
something for you to think about!

xv
Introduction

B
 ook Layout
We split this book into nine separate chapters that each highlight a specific area, or
feature, of Big Data Clusters:

–– Chapter 1: “What Are Big Data Clusters?” In this chapter we will

describe a high-level overview of SQL Server Big Data Clusters and
their various use cases.

–– Chapter 2: “Big Data Cluster Architecture.” We will go into more

depth about what makes up a Big Data Cluster in this chapter,
describing the various logical areas inside a Big Data Cluster and
looking at how all the different parts work together.

–– Chapter 3: “Deployment of Big Data Clusters.” This chapter will walk

you through the first steps of deploying a Big Data Cluster using an
on-premises or cloud environment and describe how to connect to
your cluster and finally what management options are available to
manage and monitor your Big Data Cluster.

–– Chapter 4: “Loading Data into Big Data Clusters.” This chapter will
focus on data ingression from various sources unto a Big Data
Cluster.

–– Chapter 5: “Querying Big Data Clusters Through T-SQL.” This chapter

focuses on working with external tables through PolyBase and
querying your data using T-SQL statements.

–– Chapter 6: “Working with Spark in Big Data Clusters.” While the

previous chapter focused mostly on using T-SQL to work with the
data on Big Data Clusters, this chapter puts the focus on using Spark
to perform data exploration and analysis.

–– Chapter 7: “Machine Learning on Big Data Clusters.” One of the main

features of Big Data Clusters is the ability to train, score, and opera-
tionalize machine learning models inside a single platform. In this
chapter we will focus on building and exploiting machine learning
models through SQL Server In-Database Machine Learning Services
and Spark.

xvi
 Introduction

–– Chapter 8: “Create and Consume Big Data Cluster Apps.” In the

second to last chapter of this book, we are going to take a close look at
how you can deploy and use custom applications through the Big
Data Cluster platform. These applications can range from manage-
ment tasks to providing a REST API to perform machine learning
model scoring.

–– Chapter 9: “Maintenance of Big Data Clusters.” To finish off your Big

Data Cluster experience, we’ll look at what it takes to manage and
maintain a Big Data Cluster.

xvii
CHAPTER 1

What Are Big Data

Clusters?
SQL Server 2019 Big Data Clusters – or just Big Data Clusters – are a new feature set
within SQL Server 2019 with a broad range of functionality around data virtualization,
data mart scale out, and artificial intelligence (AI).
SQL Server 2019 Big Data Clusters are only available as part of the box-product
SQL Server. This is despite Microsoft’s “cloud-first” strategy to release new features and
functionality to Azure first and eventually roll it over to the on-premises versions later
(if at all).
Major parts of Big Data Clusters run only on Linux. Let that sink in and travel back
a few years in time. If somebody had told you in early 2016 that you would be able to
run SQL Server on Linux, you probably would not have believed them. Then SQL Server
on Linux was announced, but it was only delivering a subset of what it’s “big brother” –
SQL Server on Windows – actually contained. And now we have a feature that actually
requires us to run SQL Server on Linux.
Oh, and by the way, the name is a bit misleading. Some parts of SQL Server Big Data
Clusters don’t really form a cluster – but more on that later.
Speaking of parts, Big Data Clusters is not a single feature but a huge feature set
serving a whole lot of different purposes, so it is unlikely that you will be embracing
every single piece of it. Depending on your role, specific parts may be more useful to you
than others. Over the course of this book, we will guide you through all capabilities to
allow you to pick those functions that will help you and ignore those that wouldn’t add
any value for you.

1
© Benjamin Weissman and Enrico van de Laar 2020
B. Weissman and E. van de Laar, SQL Server Big Data Clusters,
https://doi.org/10.1007/978-1-4842-5985-6_1
Chapter 1 What Are Big Data Clusters?

What Is a SQL Server 2019 Big Data Cluster Really?

SQL Server 2019 Big Data Clusters are essentially a combination of SQL Server, Apache
Spark, and the HDFS filesystem running in a Kubernetes environment. As mentioned
before, Big Data Clusters is not a single feature. Figure 1-1 categorizes the different
parts of the feature set into different groups to help you better understand what is being
provided. The overall idea is, through virtualization and scale out, SQL Server 2019
becomes your data hub for all your data, even if that data is not physically sitting in SQL
Server.

Figure 1-1. Feature overview of SQL Server 2019 Big Data Clusters

The major aspects of Big Data Clusters are shown from left to right in Figure 1-1.
You have support for data virtualization, then a managed data platform, and finally an
artificial intelligence (AI) platform. Each of these aspects is described in more detail in
the remainder of this chapter.

D
 ata Virtualization
The first feature within a SQL Server 2019 Big Data Cluster is data virtualization. Data
virtualization – unlike data integration – retains your data at the source instead of
duplicating it. Figure 1-2 illustrates this distinction between data integration and data
virtualization. The dotted rectangles in the data virtualization target represent virtual
data sources that always resolve back to a single instance of the data at the original

2
 Chapter 1 What Are Big Data Clusters?

source. In the world of Microsoft, this resolution of data to its original source is done
via a SQL Server feature named PolyBase, allowing you to virtualize all or parts of your
data mart.

Figure 1-2. Data virtualization vs. data integration

One obvious upside to data virtualization is that you get rid of redundant data as
you don’t copy it from the source but read it directly from there. Especially in cases
where you only read a big flat file once to aggregate it, there may be little to no use to
that duplicate and redundant data. Also, with PolyBase, your query is real time, whereas
integrated data will always carry some lag.
On the other hand, you can’t put indexes on an external table. Thus if you have data
that you frequently query with different workloads than on the original source, which
means that you require another indexing strategy, it might still make sense to integrate
the data rather than virtualize it. That decision may also be driven by the question on
whether you can accept the added workload to your source that would result from more
frequent reporting queries and so on.

3
Chapter 1 What Are Big Data Clusters?

Note While data virtualization solves a couple of issues that come with data
integration, it won’t be able to replace data integration. This is NOT the end of SSIS
or ETL .

Technically, PolyBase has been around since SQL Server 2016, but so far only
supported very limited types of data sources. In SQL Server 2019, PolyBase has been
greatly enhanced by support for multiple relational data sources such as SQL Server
or Oracle and NoSQL sources like MongoDB, HDFS, and all other kinds of data as we
illustrate in Figure 1-3.

Figure 1-3. PolyBase sources and capabilities in SQL Server 2019

4
 Chapter 1 What Are Big Data Clusters?

Effectively, you can query a table in another database or even on a completely

different machine as if it were a local table.
The use of PolyBase for virtualization may remind you of a linked server and there
definitely are some similarities. One big difference is that the query toward a linked
server tends to be longer and more involved than a PolyBase query. For example, here is
a typical query against a remote table:

SELECT * FROM MyOtherServer.MyDatabase.DBO.MyTable

Using PolyBase, you would write the same query more simply, as if the table were in
your local database. For example:

SELECT * FROM MyTable

PolyBase will know that the table is in a different database because you will have
created a definition in PolyBase indicating where the table can be found.
An advantage of using PolyBase is that you can move MyDatabase to another server
without having to rewrite your queries. Simply change your PolyBase data source
definition to redirect to the new data source. You can do that easily, without harming or
affecting your existing queries or views.
There are more differences between the use of linked servers and PolyBase. Table 1-1
describes some that you should be aware of.

Table 1-1. Comparison of linked servers and PolyBase

Linked Server PolyBase

− Instance scoped − Database scoped

− OLEDB providers − ODBC drivers
− Read/write and pass-through statements − Read-only operations
− Single-threaded − Queries can be scaled out
− Separate configuration needed for each − No separate configuration needed for Always On
instance in Always On Availability Group Availability Group

5
Chapter 1 What Are Big Data Clusters?

Outsource Your Data

You may have heard of “Stretch Database,”1 a feature introduced in SQL Server 2016,
which allows you to offload parts of your data to Azure. The idea is to use the feature for
“cold data” – meaning data that you don’t access as frequently because it’s either old (but
still needed for some queries) or simply for business areas that require less attention.
The rationale behind cold data is that it should be cheaper to store that data in Azure
than on premise. Unfortunately, the service may not be right for everyone as even its
entry tier provides significant storage performance which obviously comes at a cost.
With PolyBase, you can now, for example, offload data to an Azure SQL Database
and build your own very low-level outsourcing functionality.

Reduce Data Redundancy and Development Time

Besides offloading data, the reason to virtualize it instead of integrating it is obviously
the potentially tremendous reduction of data redundancy. As data virtualization keeps
the data at its original source and the data is therefore not persisted at the destination,
you basically cut your storage needs in half compared to a traditional ETL-based staging
process.

Note Our “cut in half” assertion may not be super accurate as you may not
have staged the full dataset anyway (reducing the savings) or you may have used
different datatypes (potentially even increasing the savings even more).

Think of this: You want to track the number of page requests on your website per
hour which is logging to text files. In a traditional environment, you would have written
a SQL Server Integration Services (SSIS) package to load the text file into a table, then
run a query on it to group the data, and then store or use its result. In this then new
virtualization approach, you would still run the query to group the data but you’d run
it right on your flat file, saving the time it would have taken to develop the SSIS package
and also the storage for the staging table holding the log data which would otherwise
have coexisted in the file as well as the staging table in SQL Server.

1
https://azure.microsoft.com/en-us/pricing/details/sql-server-stretch-database/

6
 Chapter 1 What Are Big Data Clusters?

A Combined Data Platform Environment

One of the big use cases of SQL Server Big Data Clusters is the ability to create an
environment that stores, manages, and analyzes data in different formats, types, and
sizes. Most notably, you get the ability to store both relational data inside the SQL Server
component and nonrelational data inside the HDFS storage subsystem. Using Big Data
Clusters allows you to create a data lake environment that can answer all your data needs
without a huge layer of complexity that comes with managing, updating, and configuring
various parts that make up a data lake.
Big Data Clusters completely take care of the installation and management of your
Big Data Cluster straight from the installation of the product. Since Big Data Clusters
is being pushed as a stand-alone product with full support from Microsoft, this means
Microsoft is going to handle updates for all the technologies that make up Big Data
Clusters through service packs and updates.
So why would you be interested in a data lake? As it turns out, many organizations
have a wide variety of data stored in different formats. In many situations, a large
portion of data comes from the use of applications that store their data inside relational
databases like SQL Server. By using a relational database, we can easily query the data
inside of it and use it for all kinds of things like dashboards, KPIs, or even machine
learning tasks to predict future sales, for instance.
A relational database must follow a number of rules, and one of the most important
of those rules is that a relational database always stores data in a schema-on-write
manner. This means that if you want to insert data into a relational database, you have
to make sure the data complies to the structure of the table being written to. Figure 1-4
illustrates schema-on-write.
For instance, a table with the columns OrderID, OrderCustomer, and
“OrderAmount” dictates that data you are inserting into that table will also need to
contain those same columns. This means that when you want to write a new row in this
table, you will have to define an OrderID, OrderCustomer, and OrderAmount for the
insert to be successful. There is no room for adding additional columns on the fly, and in
many cases, the data you are inserting needs to be the same datatype as specified in the
table (for inside integers for numbers and strings for text).

7
Chapter 1 What Are Big Data Clusters?

Figure 1-4. Scheme-on-write

Now in many situations the schema-on-write approach is perfectly fine. You make
sure all your data is formatted in the way the relational databases expect it to be, and
you can store all your data inside of it. But what happens when you decide to add new
datasets that do not necessarily have a fixed schema? Or, you want to process data that
is very large (multiple terabytes) in terms of size? In those situations, it is frequently
advised to look for another technology to store and process your data since a relational
database has difficulties handling data with those characteristics.
Solutions like Hadoop and HDFS were created to solve some of the limitations
around relational databases. Big Data platforms are able to process large volumes of data
in a distributed manner by spreading the data across different machines (called nodes)
that make up a cluster architecture. Using a technology like Hadoop, or as we will use in
this book Spark, allows you to store and process data in any format. This means we can
store huge CSV (comma-separated values) files, video files, Word documents, PDFs, or
whatever we please without having to worry about complying to a predefined schema
like we’d have to when storing data inside a relational database.
Apache’s Spark technology makes sure our data is cut up into smaller blocks and
stored on the filesystem of the nodes that make up a Spark cluster. We only have to worry
about the schema when we are going to read in and process the data, something that
is called schema-on-read. When we load in our CSV file to check its contents, we have
to define what type of data it is and, in the case of a CSV file, what the columns are of
the data. Specifying these details on read allows us a lot of flexibility when dealing with
this data, since we can add or remove columns or transform datatypes without having
to worry about a schema before we write the data back again. Because a technology
like Spark has a distributed architecture, we can perform all these data manipulation
and querying steps very quickly on large datasets, something we are explaining in more
detail in Chapter 2.
What you see in the real world is that in many situations organizations have both
relational databases and a Hadoop/Spark cluster to store and process their data. These
solutions are implemented separately from each other and, in many cases, do not “talk”
8
 Chapter 1 What Are Big Data Clusters?

to each other. Is the data relational? Store it in the database! Is it nonrelational like CSV,
IoT data, or other formats? Throw it on the Hadoop/Spark cluster! One reason why we
are so excited over the release of SQL Server Big Data Clusters is that it combines both
these solutions into a single product, a product that contains both the capabilities of a
Spark cluster together with SQL Server. And while you still must choose whether you
are going to store something directly in the SQL Server database or store it on the HDFS
filesystem, you can always access it from both technologies! Want to combine relational
data with a CSV file that is stored on HDFS? No problem, using data virtualization we
described earlier in this chapter, you can read the contents from the CSV file from HDFS
and merge it with your relational data using a T-SQL query producing a single result!
In this sense, SQL Server Big Data Clusters are made up from technologies that
complement each other very well, allowing you to bridge the gap on how limited you
are in processing data based on the manner in which it is stored. Big Data Clusters
ultimately let you create a scalable and flexible data lake environment in which you
can store and process data in any format, shape, or size, even allowing you to choose
between processing the data using SQL Server or Spark, whichever you prefer for the
tasks you want to perform.
The Big Data Cluster architecture will also be able to optimize performance in terms
of data analytics. Having all data you require stored inside a single cluster, whether it is
relational or not, means that you can access it immediately whenever you require it. You
avoid data movement across different systems or networks, which is a huge advantage in
a world where we are constantly trying to find solutions to analyze data faster and faster.
If you ask us what the ultimate advantage of SQL Server Big Data Clusters is, we
firmly believe it is the ability to store, process, and analyze data in any shape, size, or type
inside a single solution.

Centralized AI Platform
As we described in the preceding section, SQL Server Big Data Clusters allow you to
create a data lake environment that can handle all types and formats of data. Next to
having huge advantages when processing, it naturally also has immense advantages
when dealing with advanced analytics like machine learning. Since all your data is
essentially stored in one place, you can perform tasks like machine learning model
training on all the data that is available on the Big Data Cluster, instead of having to
gather data from multiple systems across your organization.

9
Chapter 1 What Are Big Data Clusters?

By combining SQL Server and Spark, we also have multiple options available when
working with machine learning. We can choose to train and score machine learning
models through Spark directly by accessing data that is stored on the HDFS filesystem, or
use the In-Database Machine Learning Services available to us through SQL Server. Both
these options allow a wide variety in languages and libraries you, or your data science
team, can use, for instance, R, Python, and Java for SQL Server Machine Learning
Services, or PySpark and Scala when running your machine learning workload through
the Spark cluster.
In terms of use cases, Big Data Clusters can handle just about any machine learning
process, from handling real-time scoring to using GPUs in combination with TensorFlow
to optimize the handling of CPU-intensive workloads or, for instance, perform image
classification tasks.

10
CHAPTER 2

Big Data Cluster

Architecture
SQL Server Big Data Clusters are made up from a variety of technologies all working together
to create a centralized, distributed data environment. In this chapter, we are going to look at
the various technologies that make up Big Data Clusters through two different views.
First, we are evaluating the more-or-less physical architecture of Big Data Clusters.
We are going to explore the use of containers, the Linux operating system, Spark, and the
HDFS storage subsystem that make up the storage layer of Big Data Clusters.
In the second part of this chapter, we are going to look at the logical architecture
which is made up of four different logical areas. These areas combine several
technologies to provide a specific function, or role(s), inside the Big Data Cluster.

Physical Big Data Cluster Infrastructure

The physical infrastructure of Big Data Clusters is made up from containers on which
you deploy the major software components. These major components are SQL Server on
Linux, Apache Spark, and the HDFS filesystem. The following is an introduction to these
infrastructure elements, beginning with containers and moving through the others to
provide you with the big picture of how the components fit together.

C
 ontainers
A container is a kind of stand-alone package that contains everything you need to run an
application in an isolated or sandbox environment. Containers are frequently compared
to virtual machines (VMs) because of the virtualization layers that are present in both
solutions. However, containers provide far more flexibility than virtual machines. A
notable area of increased flexibility is the area of portability.

11
© Benjamin Weissman and Enrico van de Laar 2020
B. Weissman and E. van de Laar, SQL Server Big Data Clusters,
https://doi.org/10.1007/978-1-4842-5985-6_2
Chapter 2 Big Data Cluster Architecture

One of the main advantages of using containers is that they avoid the
implementation of an operating system inside the container. Virtual machines require
the installation of their own operating system inside each virtual machine, whereas with
containers, the operating system of the host on which the containers are being run is
used by each container (through isolated processes). Tools like Docker enable multiple
operating systems on a single host machine by running a virtual machine that becomes
the host for your containers, allowing you to run a Linux container on Windows, for
example.
You can immediately see an advantage here: when running several virtual machines,
you also have an additional workload of maintaining the operating system on each
virtual machine with patches, configuring it, and making sure everything is running
the way it is supposed to be. With containers, you do not have those additional levels
of management. Instead, you maintain one copy of the operating system that is shared
among all containers.
Another advantage for containers over virtual machines is that containers can be
defined as a form of “infrastructure-as-code.” This means you can script out the entire
creation of a container inside a build file or image. This means that when you deploy
multiple containers with the same image or build file, they are 100% identical. Ensuring
100% identical deployment is something that can be very challenging when using virtual
machines, but is easily done using containers.
Figure 2-1 shows some differences between containers and virtual machines around
resource allocation and isolation. You can see how containers reduce the need for
multiple guest operating systems.

Figure 2-1. Virtual machine vs. containers

12
 Chapter 2 Big Data Cluster Architecture

A final advantage of containers we would like to mention (there are many more to
name, however, that would go beyond the scope of this book) is that containers can
be deployed as “stateless” applications. Essentially this means that containers won’t
change, and they do not store data inside themselves.
Consider, for instance, a situation in which you have a number of application
services deployed using containers. In this situation, each of the containers would run
the application in the exact same manner and state as the other containers in your
infrastructure. When one container crashes, it is easy to deploy a new container with
the same build file filling in the role of the crashed container, since no data inside the
containers is stored or changed for the time they are running.
The storage of your application data could be handled by other roles in your
infrastructure, for instance, a SQL Server that holds the data that is being processed by
your application containers, or, as a different example, a file share that stores the data
that is being used by the applications inside your containers. Also, when you have a
new software build available for your application servers, you can easily create a new
container image or build file, map that image or build file to your application containers,
and switch between build versions practically on the fly.
SQL Server Big Data Clusters are deployed using containers to create a scalable,
consistent, and elastic environment for all the various roles and functions that are
available in Big Data Clusters. Microsoft has chosen to deploy all the containers using
Kubernetes. Kubernetes is an additional layer in the container infrastructure that acts
like an orchestrator. By using Kubernetes (or K8s as it is often called), you get several
advantages when dealing with containers. For instance, Kubernetes can automatically
deploy new containers whenever it is required from a performance perspective, or
deploy new containers whenever others fail.
Because Big Data Clusters are built on top of Kubernetes, you have some flexibility in
where you deploy Big Data Clusters. Azure has the ability to use a managed Kubernetes
Service (AKS) where you can also choose to deploy Big Data Clusters if you so want to.
Other, on-premises options are Docker or Minikube as container orchestrators. We will
take a more in-depth look at the deployment of Big Data Clusters inside AKS, Docker, or
Minikube in Chapter 3.
Using Kubernetes also introduces a couple of specific terms that we will be using
throughout this book. We’ve already discussed the idea and definition of containers.
However, Kubernetes (and also Big Data Clusters) also frequently uses another term

13
Chapter 2 Big Data Cluster Architecture

called “pods.” Kubernetes does not run containers directly; instead it wraps a container
in a structure called a pod. A pod combines one or multiple containers, storage
resources, networking configurations, and a specific configuration governing how the
container should run inside the pod.
Figure 2-2 shows a simple representation of the node – pods – container architecture
inside Kubernetes.

Figure 2-2. Representation of containers, pods, and nodes in Kubernetes

Generally, pods are used in two manners: a single container per pod or multiple
containers inside a single pod. The latter is used when you have multiple containers
that need to work together in one way or the other – for instance, when distributing a
load across various containers. Pods are also the resource managed to allocate more
system resources to containers. For example, to increase the available memory for your
containers, a change in the pod’s configuration will result in access to the added memory
for all containers inside the pod. On that note, you are mostly managing and scaling
pods instead of containers inside a Kubernetes cluster.
Pods run on Kubernetes nodes. A node is the smallest unit of computing hardware
inside the Kubernetes cluster. Most of the time, a node is a single physical or virtual
machine on which the Kubernetes cluster software is installed, but in theory every
machine/device with a CPU and memory can be a Kubernetes node. Because these
machines only function as hosts of Kubernetes pods, they can easily be replaced, added,
or removed from the Kubernetes architecture, making the underlying physical (or
virtual) machine infrastructure very flexible.

14
 Chapter 2 Big Data Cluster Architecture

SQL Server on Linux

In March 2016, Microsoft announced that the next edition of SQL Server, which turned
out to be SQL Server 2017, would be available not only on Windows operating systems
but on Linux as well – something that seemed impossible for as long as Microsoft has
been building software suddenly became a reality and, needless to say, the entire IT
world freaked out.
In hindsight, Microsoft had perfect timing in announcing the strategic decision to
make one of its flagship products available on Linux. The incredible adaptation of new
technologies concerning containers, which we discussed in the previous section, was
mostly based on Linux distributions. We believe that without the capability’s containers,
and thus the Linux operating system those containers provide, there would never have
been a SQL Server Big Data Cluster product.
Thankfully Microsoft pushed through on their adoption of Linux, and with the
latest SQL Server 2019 release, many of the issues that plagued the SQL Server 2017
release on Linux are now resolved and many capabilities that were possible on the
Windows version have been brought to Linux as well.
So how did Microsoft manage to run an application designed for the Windows
operating system on Linux? Did they rewrite all the code inside SQL Server to make it
run on Linux? As it turns out, things are far more complicated than a rewrite of the code
base to make it Linux compatible.
To make SQL Server run on Linux, Microsoft introduced a concept called a Platform
Abstraction Layer (or PAL for short). The idea of a PAL is to separate the code needed to
run, in this case, SQL Server from the code needed to interact with the operating system.
Because SQL Server has never run on anything other than Windows, SQL Server is full
of operating system references inside its code. This would mean that getting SQL Server
to run on Linux would end up taking enormous amounts of time because of all the
operating system dependencies.

15
 Another Random Document on
Scribd Without Any Related Topics
wandering. Now I know that I shall never win the crown of the
Saints.'
The King was silent in inconsolable despair; then Astrid drew nearer
to him.
'King Olaf,' she said, 'what thou now sayest both Hjalte and
Ingegerd have told me long ago, but I would not believe that thou
wert more than a good and brave knight and noble King. It is only
now that I have lived under thy roof that my soul has begun to fear
thee. I have felt that it was worse than death to appear before thee
with a lie upon my lips. Never have I been so terrified,' Astrid
continued, 'as when I understood that thou wast a Saint. When I
saw thee burn the chips in thine hand, when I saw sickness flee at
thy bidding, and the sword fall out of thine enemy's hand when he
met thee, I was terrified unto death when I saw that thou wast a
Saint, and I resolved to die before thou knewest that I had deceived
thee.'
King Olaf did not answer. Astrid looked up at him; she saw that his
eyes were turned towards heaven. She did not know if he had heard
her.
'Ah,' she said, 'this moment have I feared every day and every hour
since I came hither. I would have died rather than live through it.'
Olaf Haraldsson was still silent.
'King Olaf,' she said, 'I would gladly give my life for thee; I would
gladly throw myself into the gray river so that thou shouldst not live
with a lying woman at thy side. The more I saw of thy holiness the
better I understood that I must go from thee. A Saint of God cannot
have a lying bondwoman at his side.'
The King was still silent, but now Astrid raised her eyes to his face;
then she cried out, terror-stricken:
'King Olaf, thy face shines.'
Whilst Astrid spoke, God had shown King Olaf a vision. He saw all
the stars of heaven leave their appointed places, and fly like
swarming bees about the universe. But suddenly they all gathered
above his head and formed a radiant crown.
'Astrid,' said he, with trembling voice, 'God hath spoken to me. It is
true what thou sayest. I shall become a Saint of God.'
His voice trembled from emotion, and his face shone in the night.
But when Astrid saw the light that surrounded his head, she arose.
For her the last hope had faded.
'Now I will go,' she said. 'Now thou knowest whom thou art. Thou
canst never more bear me at thy side. But think gently of me.
Without joy or happiness have I lived all my life. In rags have I
gone; blows have I endured. Forgive me when I am gone. My love
has done thee no harm.'
When Astrid in silent despair crossed over the bridge, Olaf
Haraldsson awoke from his ecstasy. He hastened after her.
'Why wilt thou go?' he said. 'Why wilt thou go?'
'Must I not go from thee when thou art a Saint?' she whispered
scarcely audibly.
'Thou shalt not go. Now thou canst remain,' said King Olaf. 'Before, I
was a lowly man and must fear all sin; a poor earthly King was I, too
poor to bestow on thee my grace; but now all the glory of Heaven
has been given to me. Art thou weak? I am the Lord's knight. Dost
thou fall? I can lift thee up. God hath chosen me, Astrid. Thou canst
not harm me, but I can help thee. Ah! what am I saying? In this
hour God hath so wholly and fully shed the riches of His love in my
heart that I cannot even see thou hast done wrong.'
Gently and tenderly he lifted up the trembling form, and whilst
lovingly supporting her, who was still sobbing and who could hardly
stand upright, he and Astrid went back to the King's Castle.
 From a Swedish
Homestead

III
Old Agnete

Old Agnete
An old woman went up the mountain-path with short, tripping steps.
She was little and thin. Her face was pale and wizened, but neither
hard nor furrowed. She wore a long cloak and a quilled cap. She had
a Prayer-Book in her hand and a sprig of lavender in her
handkerchief.
She lived in a hut far up the high mountain where no trees could
grow. It was lying quite close to the edge of a broad glacier, which
sent its river of ice from the snow-clad mountain peak into the
depths of the valley. There she lived quite alone. All those who had
belonged to her were dead.
It was Sunday, and she had been to church. But whatever might be
the cause, her going there had not made her happy, but sorrowful.
The clergyman had spoken about death and the doomed, and that
had affected her. She had suddenly begun to think of how she had
heard in her childhood that many of the doomed were tormented in
the region of eternal cold on the mountain right above her dwelling.
She could remember many tales about these wanderers of the
glaciers—these indefatigable shadows which were hunted from place
to place by the icy mountain winds.
All at once she felt a great terror of the mountain, and thought that
her hut was dreadfully high up. Supposing those who moved about
invisibly there wandered down the glaciers! And she who was quite
alone! The word 'alone' gave to her thoughts a still sadder turn. She
again felt the full burden of that sorrow which never left her. She
thought how hard it was to be so far away from human beings.
'Old Agnete,' she said aloud to herself, as she had got into the habit
of doing in the lonely waste, 'you sit in your hut and spin, and spin.
You work and toil all the hours of the day so as not to perish from
hunger. But is there anyone to whom you give any pleasure by being
alive? Is there anyone, old Agnete? If any of your own were living
——Yes, then, perhaps, if you lived nearer the village, you might be
of some use to somebody. Poor as you are, you could neither take
dog nor cat home to you, but you could probably now and then give
a beggar shelter. You ought not to live so far away from the
highroad, old Agnete. If you could only once in a while give a thirsty
wayfarer a drink, then you would know that it was of some use your
being alive.'
She sighed, and said to herself that not even the peasant women
who gave her flax to spin would mourn her death. She had certainly
striven to do her work honestly and well, but no doubt there were
many who could have done it better. She began to cry bitterly, when
the thought struck her that his reverence, who had seen her sitting
in the same place in church for so many, many years, would perhaps
think it a matter of perfect indifference whether she was dead or
not.
'It is as if I were dead,' she said. 'No one asks after me. I would just
as well lie down and die. I am already frozen to death from cold and
loneliness. I am frozen to the core of the heart, I am indeed. Ah me!
ah me!' she said, now she had been set a-thinking; 'if there were
only someone who really needed me, there might still be a little
warmth left in old Agnete. But I cannot knit stockings for the
mountain goats, or make the beds for the marmots, can I? I tell
Thee,' she said, stretching our her hands towards heaven,
'something Thou must give me to do, or I shall lay me down and
die.'
At the same moment a tall, stern monk came towards her. He
walked by her side because he saw that she was sorrowful, and she
told him about her troubles. She said that her heart was nearly
frozen to death, and that she would become like one of the
wanderers on the glacier if God did not give her something to live
for.
'God will assuredly do that,' said the monk.
'Do you not see that God is powerless here?' old Agnete said. 'Here
there is nothing but an empty, barren waste.'
They went higher and higher towards the snow mountains. The
moss spread itself softly over the stones; the Alpine herbs, with their
velvety leaves, grew along the pathway; the mountain, with its rifts
and precipices, its glaciers and snow-drifts, towered above them,
weighing them down. Then the monk discovered old Agnete's hut,
right below the glacier.
'Oh,' he said, 'is it there you live? Then you are not alone there; you
have company enough. Only look!'
The monk put his thumb and first finger together, held them before
old Agnete's left eye, and bade her look through them towards the
mountain. But old Agnete shuddered and closed her eyes.
'If there is anything to see up there, then I will not look on any
account,' she said. 'The Lord preserve us! it is bad enough without
that.'
'Good-bye, then,' said the monk; 'it is not certain that you will be
permitted to see such a thing a second time.'
Old Agnete grew curious; she opened her eyes and looked towards
the glacier. At first she saw nothing remarkable, but soon she began
to discern things moving about. What she had taken to be mist and
vapour, or bluish-white shadows on the ice, were multitudes of
doomed souls, tormented in the eternal cold.
Poor old Agnete trembled like an aspen leaf. Everything was just as
she had heard it described in days gone by. The dead wandered
about there in endless anguish and pain. Most of them were
shrouded in something long and white, but all had their faces and
their hands bared.
They could not be counted, there was such a multitude. The longer
she looked, the more there appeared. Some walked proud and erect,
others seemed to dance over the glacier; but she saw that they all
cut their feet on the sharp and jagged edges of the ice.
It was just as she had been told. She saw how they constantly
huddled close together, as if to warm themselves, but immediately
drew back again, terrified by the deathly cold which emanated from
their bodies.
It was as if the cold of the mountain came from them, as if it were
they who prevented the snow from melting and made the mist so
piercingly cold.
They were not all moving; some stood in icy stoniness, and it looked
as if they had been standing thus for years, for ice and snow had
gathered around them so that only the upper portion of their bodies
could be seen.
The longer the little old woman gazed the quieter she grew. Fear left
her, and she was only filled with sorrow for all these tormented
beings. There was no abatement in their pain, no rest for their torn
feet, hurrying over ice sharp as edged steel. And how cold they
were! how they shivered! how their teeth chattered from cold! Those
who were petrified and those who could move, all suffered alike
from the snarling, biting, unbearable cold.
There were many young men and women; but there was no youth in
their faces, blue with cold. It looked as if they were playing, but all
joy was dead. They shivered, and were huddled up like old people.
But those who made the deepest impression on her were those
frozen fast in the hard glacier, and those who were hanging from the
mountain-side like great icicles.
Then the monk removed his hand, and old Agnete saw only the
barren, empty glaciers. Here and there were ice-mounds, but they
did not surround any petrified ghosts. The blue light on the glacier
did not proceed from frozen bodies; the wind chased the snowflakes
before it, but not any ghosts.
Still old Agnete was certain that she had really seen all this, and she
asked the monk:
'Is it permitted to do anything for these poor doomed ones?'
He answered:
'When has God forbidden Love to do good or Mercy to solace?'
Then the monk went his way, and old Agnete went to her hut and
thought it all over. The whole evening she pondered how she could
help the doomed who were wandering on the glaciers. For the first
time in many years she had been too busy to think of her loneliness.
Next morning she again went down to the village. She smiled, and
was well content. Old age was no longer so heavy a burden. 'The
dead,' she said to herself, 'do not care so much about red cheeks
and light steps. They only want one to think of them with a little
warmth. But young people do not trouble to do that. Oh no, oh no.
How should the dead protect themselves from the terrible coldness
of death did not old people open their hearts to them?
When she came to the village shop she bought a large package of
candles, and from a peasant she ordered a great load of firewood;
but in order to pay for it she had to take in twice as much spinning
as usual.
Towards evening, when she got home again, she said many prayers,
and tried to keep up her courage by singing hymns. But her courage
sank more and more. All the same, she did what she had made up
her mind to do.
She moved her bed into the inner room of her hut. In the front room
she made a big fire and lighted it. In the window she placed two
candles, and left the outer door wide open. Then she went to bed.
She lay in the darkness and listened.
Yes, there certainly was a step. It was as if someone had come
gliding down the glacier. It came heavily, moaning. It crept round
the hut as if it dared not come in. Close to the wall it stood and
shivered.
Old Agnete could not bear it any longer. She sprang out of bed, went
into the outer room and closed the door. It was too much; flesh and
blood could not stand it.
Outside the hut she heard deep sighs and dragging steps, as of sore,
wounded feet. They dragged themselves away further and further
up the icy glacier. Now and again she also heard sobs; but soon
everything was quiet.
Then old Agnete was beside herself with anxiety. 'You are a coward,
you silly old thing,' she said. 'Both the fire and the lights, which cost
so much, are burning out. Shall it all have been done in vain
because you are such a miserable coward?' And when she had said
this she got out of bed again, crying from fear, with chattering teeth,
and shivering all over; but into the other room she went, and the
door she opened.
Again she lay and waited. Now she was no longer frightened that
they should come. She was only afraid lest she had scared them
away, and that they dared not come back.
And as she lay there in the darkness she began to call just as she
used to do in her young days when she was tending the sheep.
'My little white lambs, my lambs in the mountains, come, come!
Come down from rift and precipice, my little white lambs!'
Then it seemed as if a cold wind from the mountain came rushing
into the room. She heard neither step nor sob, only gusts of wind
that came rushing along the walls of the hut into the room. And it
sounded as if someone were continually saying:
'Hush, hush! Don't frighten her! don't frighten her! don't frighten
her!'
She had a feeling as if the outside room was so overcrowded that
they were being crushed against the walls, and that the walls were
giving way. Sometimes it seemed as if they would lift the roof in
order to gain more room. But the whole time there were whispers:
'Hush, hush! Don't frighten her! don't frighten her!'
Then old Agnete felt happy and peaceful. She folded her hands and
fell asleep. In the morning it seemed as if the whole had been a
dream. Everything looked as usual in the outer room; the fire had
burnt out, and so had the candles. There was not a vestige of tallow
left in the candlesticks.
As long as old Agnete lived she continued to do this. She spun and
worked so that she could keep her fire burning every night. And she
was happy because someone needed her.
Then one Sunday she was not in her usual seat in the church. Two
peasants went up to her hut to see if there was anything the matter.
She was already dead, and they carried her body down to the village
to bury it.
When, the following Sunday, her funeral took place, just before
Mass, there were but few who followed, neither did one see grief on
any face. But suddenly, just as the coffin was being lowered into the
grave, a tall, stern monk came into the churchyard, and he stood still
and pointed to the snow-clad mountains. Then they saw the whole
mountain-ridge shining in a red light as if lighted with joy, and round
it wound a procession of small yellow flames, looking like burning
candles. And these flames numbered as many as the candles which
old Agnete had burned for the doomed. Then people said: 'Praise
the Lord! She whom no one mourns here below has all the same
found friends in the solitude above.'
 From a Swedish
Homestead

IV
The Fisherman's Ring

The Fisherman's Ring

During the reign of the Doge Gradenigos there lived in Venice an old
fisherman, Cecco by name. He had been an unusually strong man,
and was still very strong for his age, but lately he had given up work
and left it to his two sons to provide for him. He was very proud of
his sons, and he loved them—ah, signor, how he loved them!
Fate had so ordered it that their bringing up had been almost
entirely left to him. Their mother had died early, and so Cecco had to
take care of them. He had looked after their clothes and cooked
their food; he had sat in the boat with needle and cotton and
mended and darned. He had not cared in the least that people had
laughed at him on that account. He had also, quite alone, taught
them all it was necessary for them to know. He had made a couple
of able fishermen of them, and taught them to honour God and San
Marco.
'Always remember,' he said to them, 'that Venice will never be able
to stand in her own strength. Look at her! Has she not been built on
the waves? Look at the low islands close to land, where the sea
plays amongst the seaweed. You would not venture to tread upon
them, and yet it is upon such foundation that the whole city rests.
And do you not know that the north wind has strength enough to
throw both churches and palaces into the sea? Do you not know that
we have such powerful enemies, that all the princes in Christendom
cannot vanquish them? Therefore you must always pray to San
Marco, for in his strong hands rests the chains which hold Venice
suspended over the depths of the sea.'
And in the evening, when the moon shed its light over Venice,
greenish-blue from the sea-mist; when they quietly glided up the
Canale Grande and the gondolas they met were full of singers; when
the palaces shone in their white splendour, and thousands of lights
mirrored themselves in the dark waters—then he always reminded
them that they must thank San Marco for life and happiness.
But oh, signor! he did not forget him in the daytime either. When
they returned from fishing and glided over the water of the lagoons,
light-blue and golden; when the city lay before them, swimming on
the waves; when the great ships passed in and out of the harbour,
and the palace of the Doges shone like a huge jewel-casket, holding
all the world's treasure—then he never forgot to tell them that all
these things were the gift of San Marco, and that they would all
vanish if a single Venetian were ungrateful enough to give up
believing in and adoring him.
Then, one day, the sons went out fishing on the open sea, outside
Lido. They were in company with several others, had a splendid
vessel, and intended being away several days. The weather was fine,
and they hoped for a goodly haul.
They left the Rialto, the large island where the city proper lies, one
early morning, and as they passed through the lagoons they saw all
the islands which, like fortifications, protect Venice against the sea,
appear through the mist of the morning. There were La Gindecca
and San Giorgio on the right, and San Michele, Muracco and San
Lazzaro on the left. Then island followed upon island in a large
circle, right on to the long Lido lying straight before them, and
forming, as it were, the clasp of this string of pearls. And beyond
Lido was the wide, infinite sea.
When they were well at sea, some of them got into a small boat and
rowed out to set their nets. It was still fine weather, although the
waves were higher here than inside the islands. None of them,
however, dreamt of any danger. They had a good boat and were
experienced men. But soon those left on the vessel saw that the sea
and the sky suddenly grew darker in the north. They understood
that a storm was coming on, and they at once shouted to their
comrades, but they were already too far away to hear them.
The wind first reached the small boat. When the fishermen suddenly
saw the waves rise around them, as herds of cattle on a large plain
arise in the morning, one of the men in the boat stood up and
beckoned to his comrades, but the same moment he fell backwards
into the sea. Immediately afterwards a wave came which raised the
boat on her bows, and one could see how the men, as it were, were
shaken from off their seats and flung into the sea. It only lasted a
moment, and everything had disappeared. Then the boat again
appeared, keel upwards. The men in the vessel tried to reach the
spot, but could not tack against the wind.
It was a terrific storm which came rushing over the sea, and soon
the fishermen in the vessel had their work set to save themselves.
They succeeded in getting home safely, however, and brought with
them the news of the disaster. It was Cecco's two sons and three
others who had perished.
Ah me! how strangely things come about! The same morning Cecco
had gone down to the Rialto to the fish-market. He went about
amongst the stands and strutted about like a fine gentleman
because he had no need to work. He even invited a couple of old
Lido fishermen to an asteri and stood them a beaker of wine. He
grew very important as he sat there and bragged and boasted about
his sons. His spirits rose high, and he took out the zecchine—the one
the Doge had given him when he had saved a child from drowning in
Canale Grande. He was very proud of this large gold coin, carried it
always about him, and showed it to people whenever there was an
opportunity.
Suddenly a man entered the asteri and began to tell about the
disaster, without noticing that Cecco was sitting there. But he had
not been speaking long before Cecco threw himself over him and
seized him by the throat.
'You do not dare to tell me that they are dead!' he shrieked—'not my
sons!'
The man succeeded in getting away from him, but Cecco for a long
time went on as if he were out of his mind. People heard him shout
and groan; they crowded into the asteri—as many as it could hold—
and stood round him in a circle as if he were a juggler.
Cecco sat on the floor and moaned. He hit the hard stone floor with
his fist, and said over and over again:
'It is San Marco, San Marco, San Marco!'
'Cecco, you have taken leave of your senses from grief,' they said to
him.
'I knew it would happen on the open sea,' Cecco said; 'outside Lido
and Malamocco, there, I knew it would happen. There San Marco
would take them. He bore them a grudge. I have feared it, boy. Yes,'
he said, without hearing what they said to quiet him, 'they once
laughed at him, once when we were lying outside Lido. He has not
forgotten it; he will not stand being laughed at.'
He looked with confused glances at the bystanders, as if to seek
help.
'Look here, Beppo from Malamocca,' he said, stretching out his hand
towards a big fisherman, 'don't you believe it was San Marco?'
'Don't imagine any such thing, Cecco.'
'Now you shall hear, Beppo, how it happened. You see, we were
lying out at sea, and to while away the time I told them how San
Marco had come to Venice. The evangelist San Marco was first
buried in a beautiful cathedral at Alexandria in Egypt. But the town
got into the possession of unbelievers, and one day the Khalifa
ordered that they should build him a magnificent palace at
Alexandria, and take some columns from the Christian churches for
its decoration. But just at that time there were two Venetian
merchants at Alexandria who had ten heavily-laden vessels lying in
the harbour. When these men entered the church where San Marco
was buried and heard the command of the Khalifa, they said to the
sorrowful priests: "The precious body which you have in your church
may be desecrated by the Saracens. Give it to us; we will honour it,
for San Marco was the first to preach on the Lagoon, and the Doge
will reward you." And the priests gave their consent, and in order
that the Christians of Alexandria should not object, the body of
another holy man was placed in the Evangelist's coffin. But to
prevent the Saracens from getting any news of the removal of the
body, it was placed at the bottom of a large chest, and above it were
packed hams and smoked bacon, which the Saracens could not
endure. So when the Custom-house officers opened the lid of the
chest, they at once hurried away. The two merchants, however,
brought San Marco safely to Venice; you know, Beppo, that this is
what they say.'
'I do, Cecco.'
'Yes; but just listen now,' and Cecco half arose, and in his fear spoke
in a low voice. 'Something terrible now happened. When I told the
boys that the holy man had been hidden underneath the bacon, they
burst out laughing. I tried to hush them, but they only laughed the
louder. Giacomo was lying on his stomach in the bows, and Pietro
sat with his legs dangling outside the boat, and they both laughed
so that it could be heard far out over the sea.'
'But, Cecco, surely two children may be allowed to laugh.'
'But don't you understand that is where they have perished to-day—
on the very spot? Or can you understand why they should have lost
their lives on that spot?'
Now they all began to talk to him and comfort him. It was his grief
which made him lose his senses. This was not like San Marco. He
would not revenge himself upon two children. Was it not natural that
when a boat was caught in a storm this would happen on the open
sea and not in the harbour?
Surely his sons had not lived in enmity with San Marco. They had
heard them shout, 'Eviva San Marco!' as eagerly as all the others,
and had he not protected them to this very day. He had never,
during the years that had passed, shown any sign of being angry
with them.
'But, Cecco,' they said, 'you will bring misfortune upon us with your
talk about San Marco. You, who are an old man and a wise man,
should know better than to raise his anger against the Venetians.
What are we without him?'
Cecco sat and looked at them bewildered.
'Then you don't believe it?'
'No one in his senses would believe such a thing.'
It looked as if they had succeeded in quieting him.
'I will also try not to believe it,' he said. He rose and walked towards
the door. 'It would be too cruel, would it not?' he said. 'They were
too handsome and too brave for anyone to hate them; I will not
believe it.'
He went home, and in the narrow street outside his door he met an
old woman, one of his neighbours.
'They are reading a Mass in the cathedral for the souls of the dead,'
she said to Cecco, and hurried away. She was afraid of him; he
looked so strange.
Cecco took his boat and made his way through the small canals
down to Riva degli Schiavoni. There was a wide view from there; he
looked towards Lido and the sea. Yes, it was a hard wind, but not a
storm by any means; there were hardly any waves. And his sons had
perished in weather like this! It was inconceivable.
He fastened his boat, and went across the Piazetta and the Market
Place into San Marco. There were many people in the church, and
they were all kneeling and praying in great fear; for it is much more
terrible for the Venetians, you know, than any other people when
there is a disaster at sea. They do not get their living from vineyards
or fields, but they are all, everyone of them, dependent on the sea.
Whenever the sea rose against any one of them they were all afraid,
and hurried to San Marco to pray to him for protection.
As soon as Cecco entered the cathedral he stopped. He thought of
how he had brought his little sons there, and taught them to pray to
San Marco. 'It is he who carries us over the sea, who opens the
gates of Byzance for us and gives us the supremacy over the islands
of the East,' he said to them. Out of gratitude for all this the
Venetians had built San Marco the most beautiful temple in the
world, and no vessel ever returned from a foreign port without
bringing a gift for San Marco.
Then they had admired the red marble walls of the cathedral and
the golden mosaic ceiling. It was as if no misfortune could befall a
city that had such a sanctuary for her patron Saint.
Cecco quickly knelt down and began to pray, the one Paternoster
after the other. It came back, he felt. He would send it away by
prayers. He would not believe anything bad about San Marco.
But it had been no storm at all. And so much was certain, that even
if the Saint had not sent the storm, he had, in any case, not done
anything to help Cecco's sons, but had allowed them to perish as if
by accident. When this thought came upon him he began to pray;
but the thought would not leave him.
And to think that San Marco had a treasury in this cathedral full of all
the glories of fairyland! To think that he had himself prayed to him
all his life, and had never rowed past the Piazetta without going into
the cathedral to invoke him!
Surely it was not by a mere accident that his sons had to-day
perished on the sea! Oh, it was miserable for the Venetians to have
no one better to depend upon! Just fancy a Saint who revenged
himself upon two children—a patron Saint who could not protect
against a gust of wind!
He stood up, and he shrugged his shoulders, and disparagingly
waved his hand when he looked towards the tomb of the Saint in the
chancel.
A verger was going about with a large chased silver-gilt dish,
collecting gifts for San Marco. He went from the one person to the
other, and also came to Cecco.
Cecco drew back as if it were the Evil One himself who handed him
the plate. Did San Marco ask for gifts from him? Did he think he
deserved gifts from him?
All at once he seized the large golden zecchine he had in his belt,
and flung it into the plate with such violence that the ring of it could
be heard all over the church. It disturbed those who were praying,
and made them turn round. And all who saw Cecco's face were
terrified; he looked as if he were possessed of evil spirits.
Cecco immediately left the church, and at first felt it as a great relief
that he had been revenged upon the Saint. He had treated him as
one treats a usurer who demands more than he is entitled to. 'Take
this too,' one says, and throws his last gold piece in the fellow's face
so that the blood runs down over his eyes. But the usurer does not
strike again—simply stoops and picks up the zecchine. So, too, had
San Marco done. He had accepted Cecco's zecchine, having first
robbed him of his sons. Cecco had made him accept a gift which had
been tendered with such bitter hatred. Would an honourable man
have put up with such treatment? But San Marco was a coward—
both cowardly and revengeful. But he was not likely to revenge
himself upon Cecco. He was, no doubt, pleased and thankful he had
got the zecchine. He simply accepted it and pretended that it had
been given as piously as could be.
When Cecco stood at the entrance, two vergers quickly passed him.
'It rises—it rises terribly!' the one said.
'What rises?' asked Cecco.
'The water in the crypt. It has risen a foot in the last two or three
minutes.'
When Cecco went down the steps, he saw a small pool of water on
the Market Place close to the bottom step. It was sea-water, which
had splashed up from the Piazetta. He was surprised that the sea
had risen so high, and he hurried down to the Riva, where his boat
lay. Everything was as he had left it, only the water had risen
considerably. It came rolling in broad waves through the five sea-
gates; but the wind was not very strong. At the Riva there were
already pools of sea-water, and the canals rose so that the doors in
the houses facing the water had to be closed. The sky was all gray
like the sea.
It never struck Cecco that it might grow into a serious storm. He
would not believe any such thing. San Marco had allowed his sons to
perish without cause. He felt sure this was no real storm. He would
just like to see if it would be a storm, and he sat down beside his
boat and waited.
Then suddenly rifts appeared in the dull-gray clouds which covered
the sky. The clouds were torn asunder and flung aside, and large
storm-clouds came rushing, black like warships, and from them
scourging rain and hail fell upon the city. And something like quite a
new sea came surging in from Lido. Ah, signor! they were not the
swan-necked waves you have seen out there, the waves that bend
their transparent necks and hasten towards the shore, and which,
when they are pitilessly repulsed, float away again with their white
foam-hair dispersed over the surface of the sea. These were dark
waves, chasing each other in furious rage, and over their tops the
bitter froth of the sea was whipped into mist.
The wind was now so strong that the seagulls could no longer
continue their quiet flight, but, shrieking, were thrust from their
course. Cecco soon saw them with much trouble making their way
towards the sea, so as not to be caught by the storm and flung
against the walls. Hundreds of pigeons on San Marco's square flew
up, beating their wings, so that it sounded like a new storm, and hid
themselves away in all the nooks and corners of the church roof.
But it was not the birds alone that were frightened by the storm. A
couple of gondolas had already got loose, and were thrown against
the shore, and were nearly shattered. And now all the gondoliers
came rushing to pull their boats into the boathouses, or place them
in shelter in the small canals.
The sailors on the ships lying in the harbour worked with the anchor-
chains to make the vessels fast, in order to prevent them drifting on
to the shore. They took down the clothes hanging up to dry, pulled
their long caps well over their foreheads, and began to collect all the
loose articles lying about in order to bring them below deck. Outside
Canale Grande a whole fishing-fleet came hurrying home. All the
people from Lido and Malamocco who had sold their goods at the
Rialto were rushing homewards, before the storm grew too violent.
Cecco laughed when he saw the fishermen bending over their oars
and straining themselves as if they were fleeing from death itself.
Could they not see that it was only a gust of wind? They could very
well have remained and given the Venetian women time to buy all
their cattle, fish, and crabs.
He was certainly not going to pull his boat into shelter, although the
storm was now violent enough for any ordinary man to have taken
notice of it. The floating bridges were lifted up high and cast on to
the shore, whilst the washerwomen hurried home shrieking. The
broad-brimmed hats of the signors were blown off into the canals,
from whence the street-boys fished them out with great glee. Sails
were torn from the masts, and fluttered in the air with a cracking
sound; children were knocked down by the strong wind; and the
clothes hanging on the lines in the narrow streets were torn to rags
and carried far away.
Cecco laughed at the storm—a storm which drove the birds away,
and played all sorts of pranks in the street, like a boy. But, all the
same, he pulled his boat under one of the arches of the bridge. One
could really not allow what that wind might take it into its head to
do.
In the evening Cecco thought that it would have been fun to have
been out at sea. It would have been splendid sailing with such a
fresh wind. But on shore it was unpleasant. Chimneys were blown
down; the roofs of the boathouses were lifted right off; it rained tiles
from the houses into the canals; the wind shook the doors and the
window-shutters, rushed in under the open loggias of the palaces
and tore off the decorations.
Cecco held out bravely, but he did not go home to bed. He could not
take the boat home with him, so it was better to remain and look
after it. But when anyone went by and said that it was terrible
weather he would not admit it. He had experienced very different
weather in his young days.
'Storm!' he said to himself—'call this a storm? And they think,
perhaps, that it began the same moment I threw the zecchine to
San Marco. As if he can command a real storm!'
When night came the wind and the sea grew still more violent, so
that Venice trembled in her foundations. Doge Gradenigo and the
Gentlemen of the High Council went in the darkness of the night to
San Marco to pray for the city. Torch-bearers went before them, and
the flames were spread out by the wind, so that they lay flat, like
pennants. The wind tore the Doge's heavy brocade gown, so that
two men were obliged to hold it.
Cecco thought this was the most remarkable thing he had ever seen
—Doge Gradenigo going himself to the cathedral on account of this
bit of a wind! What would those people have done if there had been
a real storm?
The waves beat incessantly against the bulwarks. In the darkness of
the night it was as if white-headed wresters sprang up from the
deep, and with teeth and claws clung fast to the piles to tear them
loose from the shore. Cecco fancied he could hear their angry snorts
when they were hurled back again. But he shuddered when he
heard them come again and again, and tear in the bulwarks.
It seemed to him that the storm was far more terrible in the night.
He heard shouts in the air, and that was not the wind. Sometimes
black clouds came drifting like a whole row of heavy galleys, and it
seemed as if they advanced to make an assault on the city. Then he
heard distinctly someone speaking in one of the riven clouds over his
head.
'Things look bad for Venice now,' it said from the one cloud. 'Soon
our brothers the evil spirits will come and overthrow the city.'
'I am afraid San Marco will not allow it to happen,' came as a
response from the other cloud.
'San Marco has been knocked down by a Venetian, so he lies
powerless, and cannot help anyone,' said the first.
The storm carried the words down to old Cecco, and from that
moment he was on his knees, praying San Marco for grace and
forgiveness. For the evil spirits had spoken the truth. It did indeed
look bad for Venice. The fair Queen of the Isles was near
destruction. A Venetian had mocked San Marco, and therefore
Venice was in danger of being carried away by the sea. There would
be no more moonlight sails or her sea and in her canals, and no
more barcaroles would be heard from her black gondolas. The sea
would wash over the golden-haired signoras, over the proud palaces,
over San Marco, resplendent with gold.
If there was no one to protect these islands, they were doomed to
destruction. Before San Marco came to Venice it had often happened
that large portions of them had been washed away by the waves.
At early dawn San Marco's Church bells began to ring. People crept
to the church, their clothes being nearly torn off them.
The storm went on increasing. The priests had resolved to go out
and adjure the storm and the sea. The main doors of the cathedral
were opened, and the long procession streamed out of the church.
Foremost the cross was carried, then came the choir-boys with wax
candles, and last in the procession were carried the banner of San
Marco and the Sacred Host.
But the storm did not allow itself to be cowed; on the contrary, it
was as if it wished for nothing better to play with. It upset the choir-
boys, blew out the wax candles, and flung the baldachin, which was
carried over the Host, on to the top of the Doge's palace. It was with
the utmost trouble that they saved San Marco's banner, with the
winged lion, from being carried away.
Cecco saw all this, and stole down to his boat moaning loudly. The
whole day he lay near the shore, often wet by the waves and in
danger of being washed into the sea. The whole day he was praying
incessantly to God and San Marco. He felt that the fate of the whole
city depended upon his prayers.
There were not many people about that day, but some few went
moaning along the Riva. All spoke about the immeasurable damage
the storm had wrought. One could see the houses tumbling down on
the Murano. It was as if the whole island were under water. And also
on the Rialto one or two houses had fallen.
The storm continued the whole day with unabated violence. In the
evening a large multitude of people assembled at the Market Place
and the Piazetta, although these were nearly covered with water.
People dared not remain in their houses, which shook in their very
foundations. And the cries of those who feared disaster mingled with
the lamentations of those whom it had already overtaken. Whole
dwellings were under water; children were drowned in their cradles.
The old and the sick had been swept with the overturned houses
into the waves.
Cecco was still lying and praying to San Marco. Oh, how could the
crime of a poor fisherman be taken in such earnest? Surely it was
not his fault that the saint was so powerless! He would let the
demons take him and his boat; he deserved no better fate. But not
the whole city!—oh, God in heaven, not the whole city!
'My sons!' Cecco said to San Marco. 'What do I care about my sons
when Venice is at stake! I would willingly give a son for each tile in
danger of being blown into the canal if I could keep them in their
place at that price. Oh, San Marco, each little stone of Venice is
worth as much as a promising son.'
At times he saw terrible things. There was a large galley which had
torn itself from its moorings and now came drifting towards the
shore. It went straight against the bulwark, and struck it with the
ram's head in her bows, just as if it had been an enemy's ship. It
gave blow after blow, and the attack was so violent that the vessel
immediately sprang a leak. The water rushed in, the leak grew
larger, and the proud ship went to pieces. But the whole time one
could see the captain and two or three of the crew, who would not
leave the vessel, cling to the deck and meet death without
attempting to escape it.
The second night came, and Cecco's prayers continued to knock at
the gate of heaven.
'Let me alone suffer!' he cried. 'San Marco, it is more than a man can
bear, thus to drag others with him to destruction. Only send thy lion
and kill me; I shall not attempt to escape. Everything that thou wilt
have me give up for the city, that will I willingly sacrifice.'
Just as he had uttered these words he looked towards the Piazetta,
and he thought he could no longer see San Marco's lion on the
granite pillar. Had San Marco permitted his lion to be overthrown?
old Cecco cried. He was nearly giving up Venice.
Whilst he was lying there he saw visions and heard voices all the
time. The demons talked and moved to and fro. He heard them
wheeze like wild beasts every time they made their assaults on the
bulwarks. He did not mind them much; it was worse about Venice.
Then he heard in the air above him the beating of strong wings; this
was surely San Marco's lion flying overhead. It moved backwards
and forwards in the air; he saw and yet he did not see it. Then it
seemed to him as if it descended on Riva degli Schiavoni, where he
was lying, and prowled about there. He was on the point of jumping
into the sea from fear, but he remained sitting where he was. It was
no doubt he whom the lion sought. If that could only save Venice,
then he was quite willing to let San Marco avenge himself upon him.
Then the lion came crawling along the ground like a cat. He saw it
making ready to spring. He noticed that it beat its wings and
screwed its large carbuncle eyes together till they were only small
fiery slits.
Then old Cecco certainly did think of creeping down to his boat and
hiding himself under the arch of the bridge, but he pulled himself
together and remained where he was. The same moment a tall,
imposing figure stood by his side.
'Good-evening, Cecco,' said the man; 'take your boat and row me
across to San Giorgio Maggiore.'
'Yes, signor,' immediately replied the old fisherman.
It was as if he had awakened from a dream. The lion had
disappeared, and the man must be somebody who knew him,
although Cecco could not quite remember where he had seen him
before. He was glad to have company. The terrible heaviness and
anguish that had been over him since he had revolted against the
Saint suddenly vanished. As to rowing across to San Giorgio, he did
not for a moment think that it could be done.
'I don't believe we can even get the boat out,' he said to himself.
But there was something about the man at his side that made him
feel he must do all he possibly could to serve him; and he did
succeed in getting out the boat. He helped the stranger into the boat
and took the oars.
Cecco could not help laughing to himself.
'What are you thinking about? Don't go out further in any case,' he
said. 'Have you ever seen the like of these waves? Do tell him that it
is not within the power of man.'
But he felt as if he could not tell the stranger that it was impossible.
He was sitting there as quietly as if he were sailing to the Lido on a
summer's eve. And Cecco began to row to San Giorgio Maggiore.
It was a terrible row. Time after time the waves washed over them.
'Oh, stop him!' Cecco said under his breath; 'do stop the man who
goes to sea in such weather! Otherwise he is a sensible old
fisherman. Do stop him!'
Now the boat was up a steep mountain, and then it went down into
a valley. The foam splashed down on Cecco from the waves that
rushed past him like runaway horses, but in spite of everything he
approached San Giorgio.
'For whom are you doing all this, risking boat and life?' he said. 'You
don't even know whether he can pay you. He does not look like a
fine gentleman. He is no better dressed than you are.'
But he only said this to keep up his courage, and not to be ashamed
of his tractability. He was simply compelled to do everything the man
in the boat wanted.
'But in any case not right to San Giorgio, you foolhardy old man,' he
said. 'The wind is even worse there than at the Rialto.'
But he went there, nevertheless, and made the boat fast whilst the
stranger went on shore. He thought the wisest thing he could do
would be to slip away and leave his boat, but he did not do it. He
would rather die than deceive the stranger. He saw the latter go into
the Church of San Giorgio. Soon afterwards he returned,
accompanied by a knight in full armour.
'Row us now to San Nicolo in Lido,' said the stranger.
'Ay, ay,' Cecco thought; 'why not to Lido?' They had already, in
constant anguish and death, rowed to San Giorgio; why should they
not set out for Lido?
And Cecco was shocked at himself that he obeyed the stranger even
unto death, for he now actually steered for the Lido.
Being now three in the boat, it was still heavier work. He had no
idea how he should be able to do it. 'You might have lived many
years yet,' he said sorrowfully to himself. But the strange thing was
that he was not sorrowful, all the same. He was so glad that he
could have laughed aloud. And then he was proud that he could
make headway. 'He knows how to use his oars, does old Cecco,' he
said.
They laid-to at Lido, and the two strangers went on shore. They
walked towards San Nicolo in Lido, and soon returned accompanied
by an old Bishop, with robe and stole, crosier in hand, and mitre on
head.
'Now row out to the open sea,' said the first stranger.
Old Cecco shuddered. Should he row out to the sea, where his sons
perished? Now he had not a single cheerful word to say to himself.
He did not think so much of the storm, but of the terror it was to
have to go out to the graves of his sons. If he rowed out there, he
felt that he gave the stranger more than his life.
The three men sat silently in the boat as if they were on watch.
Cecco saw them bend forward and gaze into the night. They had
reached the gate of the sea at Lido, and the great storm-ridden sea
lay before them.
Cecco sobbed within himself. He thought of two dead bodies rolling
about in these waves. He gazed into the water for two familiar faces.
But onward the boat went. Cecco did not give in.
Then suddenly the three men rose up in the boat; and Cecco fell
upon his knees, although he still went on holding the oars. A big
ship steered straight against them.
Cecco could not quite tell whether it was a ship or only drifting mist.
The sails were large, spread out, as it were, towards the four
corners of heaven; and the hull was gigantic, but it looked as if it
were built of the lightest sea-mist. He thought he saw men on board
and heard shouting; but the crew were like deep darkness, and the
shouting was like the roar of the storm.
However it was, it was far too terrible to see the ship steer straight
upon them, and Cecco closed his eyes.
But the three men in the boat must have averted the collision, for
the boat was not upset. When Cecco looked up the ship had fled out
to sea, and loud wailings pierced the night.
He rose, trembling to row further. He felt so tired that he could
hardly hold the oars. But now there was no longer any danger. The
storm had gone down, and the waves speedily laid themselves to
rest.
'Now row us back to Venice,' said the stranger to the fisherman.
Cecco rowed the boat to Lido, where the Bishop went on shore, and
to San Giorgio, where the knight left them. The first powerful
stranger went with him all the way to the Rialto.
When they had landed at Riva degli Schiavoni he said to the
fisherman:
'When it is daylight thou shalt go to the Doge and tell him what thou
hast seen this night. Tell him that San Marco and San Giorgio and
San Nicolo have to-night fought the evil spirits that would destroy
Venice, and have put them to flight.'
'Yes, signor,' the fisherman answered, 'I will tell everything. But how
shall I speak so that the Doge will believe me?'
Then San Marco handed him a ring with a precious stone possessed
of a wonderful lustre.
'Show this to the Doge,' he said, 'then he will understand that it
brings a message from me. He knows my ring, which is kept in San
Marco's treasury in the cathedral.'
The fisherman took the ring, and kissed it reverently.
'Further, thou shalt tell the Doge,' said the holy man, 'that this is a
sign that I shall never forsake Venice. Even when the last Doge has
left Palazzo Ducali I will live and preserve Venice. Even if Venice lose
her islands in the East and the supremacy of the sea, and no Doge
ever again sets out on the Bucintoro, even then I will preserve the
city beautiful and resplendent. It shall always be rich and beloved,
always be lauded and its praises sung, always a place of joy for men
to live in. Say this, Cecco, and the Doge will not forsake thee in thine
old age.'
Then he disappeared; and soon the sun rose above the gate of the
sea at Torcello. With its first beautiful rays it shed a rosy light over
the white city and over the sea that shone in many colours. A red
glow lay over San Giorgio and San Marco, and over the whole shore,
studded with palaces. And in the lovely morning radiant Venetian
ladies came out on to the loggias and greeted with smiles the rising
day.
Venice was once again the beautiful goddess, rising from the sea in
her shell of rose-coloured pearl. Beautiful as never before, she
combed her golden hair, and threw the purple robe around her, to
begin one of her happiest days. For a transport of bliss filled her
when the old fisherman brought San Marco's ring to the Doge, and
she heard how the Saint, now, and until the end of time, would hold
his protecting hand over her.
 From a Swedish
Homestead

V
Santa Caterina of Siena

Santa Caterina of Siena

At Santa Caterina's house in Siena, on a day towards the end of
April, in the week when her fête is being celebrated, people come to
the old house in the Street of the Dyers, to the house with the pretty
loggia and with the many small chambers, which have now been
converted into chapels and sanctuaries, bringing bouquets of white
lilies; and the rooms are fragrant with incense and violets.
Walking through these rooms, one cannot help thinking that it is just
as if she were dead yesterday, as if all those who go in and out of
her home to-day had seen and known her.
But, on the other hand, no one could really think that she had died
recently, for then there would be more grief and tears, and not only
a quiet sense of loss. It is more as if a beloved daughter had been
recently married, and had left the parental home.
Look only at the nearest houses. The old walls are still decorated as
if for a fête. And in her own home garlands of flowers are still
hanging beneath the portico and loggia, green leaves are strewn on
the staircase and the doorstep, and large bouquets of flowers fill the
rooms with their scent.
She cannot possibly have been dead five hundred years. It looks
much more as if she had celebrated her marriage, and had gone
away to a country from which she would not return for many years,
perhaps never. Are not the houses decorated with nothing but red
table-cloths, red trappings, and red silken banners, and are there not
stuck red-paper roses in the dark garlands of oak-leaves? and the
hangings over the doors and the windows, are they not red with
golden fringes? Can one imagine anything more cheerful?
And notice how the old women go about in the house and examine
her small belongings. It is as if they had seen her wear that very veil
and that very shirt of hair. They inspect the room in which she lived,
and point to the bedstead and the packets of letters, and they tell
how at first she could not at all learn to write, but that it came to
her all at once without her having learnt it. And only look at her
writing—how good and distinct! And then they point to the little
bottle she used to carry at her belt, so as always to have a little
medicine at hand in case she met a sick person, and they utter a
blessing over the old lantern she held in her hand when she went
and visited the sick in the long weary nights. It is just as if they
would say: 'Dear me—dear me! that our little Caterina Benincasa
should be gone, that she will never come any more and look after us
old people!' And they kiss her picture, and take a flower from the
bouquets to keep as a remembrance.
It looks as if those who were left in the home had long ago prepared
themselves for the separation, and tried to do everything possible to
keep alive the memory of the one who had gone away. See, there
they have painted her on the wall; there is the whole of her little
history represented in every detail. There she is when she cut off her
beautiful long hair so that no man could ever fall in love with her, for
she would never marry. Oh dear—oh dear! how much ridicule and
scoffing she had suffered on that account! It is dreadful to think how
her mother tormented her and treated her like a servant, and made
her sleep on the stone floor in the hall, and would not give her any
food, all because of her being so obstinate about that hair. But what
was she to do when they continually tried to get her married—she
who would have no other bridegroom than Christ? And there she is
when she was kneeling in prayer, and her father coming into the
room without her knowing it saw a beautiful white dove hovering
over her head whilst she was praying. And there she is on that
Christmas Eve when she had gone secretly to the Madonna's altar in
order the more fully to rejoice over the birth of the Son of God, and
the beautiful Madonna leaned out of her picture and handed the
Child to her that she might be allowed to hold it for a moment in her
arms. Oh, what a joy it had been for her!
Oh dear, no; it is not at all necessary to say that our little Caterina
Benincasa is dead. One need only say that she has gone away with
the Bridegroom.
In her home one will never forget her pious ways and doings. All the
poor of Siena come and knock at her door because they know that it
is the marriage-day of the little virgin, and large piles of bread lie in
readiness for them as if she were still there. They have their pockets
and baskets filled; had she herself been there, she could not have
sent them away more heavily laden. She who had gone away had
left so great a want that one almost wonders the Bridegroom had
the heart to take her away with him.
In the small chapels which have been arranged in every corner of
the house they read Mass the whole day, and they invoke the bride
and sing hymns in her praise.
'Holy Caterina,' they say, 'on this the day of thy death, which is thine
heavenly wedding-day, pray for us!'
'Holy Caterina, thou who hadst no other love but Christ, thou who in
life wert His affianced bride, and who in death wast received by Him
in Paradise, pray for us!'
'Holy Caterina, thou radiant heavenly bride, thou most blessed of
virgins, thou whom the mother of God exalted to her Son's side,
thou who on this day wast carried by angels to the kingdom of glory,
pray for us!'

It is strange how one comes to love her, how the home and the
pictures and the love of the old and the poor seem to make her
living, and one begins to wonder how she really was, whether she
was only a saint, only a heavenly bride, and if it is true that she was
unable to love any other than Christ. And then comes to one's mind
an old story which warmed one's heart long ago, at first quite vague
and without shape, but whilst one is sitting there under the loggia in
the festively decorated home and watching the poor wander away
with their full baskets, and hearing the subdued murmur from the
chapels, the story becomes more and more distinct, and suddenly it
is vivid and clear.

Nicola Tungo was a young nobleman of Perugia, who often came to

Siena on account of the races. He soon found out how badly Siena
was governed, and often said, both at the festive gatherings of the
great and when he sat drinking in the inns, that Siena ought to rise
against the Signoria and procure other rulers.
The Signoria had not been in power for more than half a year; they
did not feel particularly firm in their office, and did not like the
Perugian stirring up the people. In order promptly to put a stop to it,
they had him imprisoned, and after a short trial he was sentenced to
death. He was placed in a cell in the Palazzo Publico whilst
preparations were being made for his execution, which was to take
place the next morning in the Market Place.
At first he was strangely affected. To-morrow he would no more
wear his green velvet doublet and his beautiful sword; he would no
more walk down the street in his cap with the ostrich-feather and
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com