This is an automated email from the ASF dual-hosted git repository.
git-site-role pushed a commit to branch asf-site
in repository https://gitbox.apache.org/repos/asf/drill-site.git
The following commit(s) were added to refs/heads/asf-site by this push:
new 81012cf71 Automatic Site Publish by Buildbot
81012cf71 is described below
commit 81012cf71613c16ddd64afd3e827b4617efd6a1d
Author: buildbot <[email protected]>
AuthorDate: Thu Mar 2 13:18:25 2023 +0000
Automatic Site Publish by Buildbot
---
.../2023-03-02-drill-1.21-announcement/index.html | 65 +++++++++++
output/blog/index.html | 5 +
output/feed.xml | 119 ++++++++++++++-------
output/index.html | 4 +-
.../2023-03-02-drill-1.21-announcement/index.html | 65 +++++++++++
output/zh/blog/index.html | 5 +
output/zh/feed.xml | 119 ++++++++++++++-------
output/zh/index.html | 4 +-
8 files changed, 302 insertions(+), 84 deletions(-)
diff --git
a/output/blog/2023/03/02/2023-03-02-drill-1.21-announcement/index.html
b/output/blog/2023/03/02/2023-03-02-drill-1.21-announcement/index.html
new file mode 100644
index 000000000..6a3135e20
--- /dev/null
+++ b/output/blog/2023/03/02/2023-03-02-drill-1.21-announcement/index.html
@@ -0,0 +1,65 @@
+<h1
id="announcing-drill-121-new-connectors-functions-and-much-better-stability">Announcing
Drill 1.21: New Connectors, Functions and Much Better Stability</h1>
+<p>The Apache Drill PMC is pleased to announce a milestone release of Apache
Drill. Since the last release of Drill the team has been hard at work quashing
bugs and making overall functionality improvements. The TL;DR includes the
following:</p>
+
+<ul>
+ <li>New connectors including Apache Iceberg, Delta Lake, Microsoft Access,
GoogleSheets, and Box</li>
+ <li>Efficient cross-cloud query capability</li>
+ <li>Greatly improved access controls to include user translation support for
all storage plugins</li>
+ <li>Greatly improved query planning and implicit casting.</li>
+ <li>New BI-focused SQL operators including <code class="language-plaintext
highlighter-rouge">PIVOT</code>, <code class="language-plaintext
highlighter-rouge">UNPIVOT</code>, <code class="language-plaintext
highlighter-rouge">EXCEPT</code> and <code class="language-plaintext
highlighter-rouge">INTERSECT</code></li>
+ <li>New functions for computing regression lines and trends.</li>
+ <li>New and updated date manipulation functions.</li>
+</ul>
+
+<p>Overall, Drill 1.21 is much more capable and stable than previous
versions.</p>
+
+<h2 id="calcite-were-back">Calcite, We’re Back!</h2>
+<p>Drill relies on another open source project, Apache Calcite for its query
planning. The query planning process is a huge part of the overall
functionality of Drill. Unfortunately, about three years ago, there were some
issues in Calcite which forced Drill to fork it and rely on that fork. As a
result, Drill was essentially stuck with a three year old query planner, but
more importantly, bugs that were fixed in Calcite, as well as new capabilities
were not finding their way into Drill.</p>
+
+<p>That is no longer the case. Drill 1.21 is now running on the latest stable
version of Calcite, version 1.33. As a result, we’ve been able to close
countless JIRA tickets of various queries failing and other random bugs that
were the result of query planning bugs.</p>
+
+<p>What this means for you as a user is that you’ll see much fewer queries
failing and better overall performance in terms of speed and stability. You’ll
see better optimizations being pushed down to JDBC data sources as well as
support for BigQuery, Athena and other JDBC data sources. We hope to keep Drill
away from Calcite forks so I hope that we will work with the Calcite community
to keep our tools in sync.</p>
+
+<h2
id="improved-implicit-casting-rules-reduce-schema-change-failures">Improved
Implicit Casting Rules Reduce Schema Change Failures</h2>
+<p>From this author’s perspective, one of the biggest improvements in Drill is
one of the least noticeable and that is the result of improved implicit
casting. One of Drill’s unique features is its ability to infer the structure,
or schema of your data. However, this can be problematic when the schema
changes. When I used to teach Drill, I used to have spend a considerable amount
of time teaching students how to cast data from one data type to another to
ensure that the queries would suc [...]
+
+<p>When using latest version of Drill, you’ll find that queries will work
without the need for much if any casting. In short, they’ll do what you expect
them to do. It’s really a high on magic functionality.</p>
+
+<h2
id="integrations-with-the-modern-and-not-so-modern-data-stack">Integrations
with the Modern and Not-so-Modern Data Stack</h2>
+<p>The new version of Drill features several new connectors and readers that
will enable users to connect to the “modern data stack”, specifically support
for Apache Iceberg and Delta Lake.</p>
+
+<h3 id="breaking-the-iceberg">Breaking the Iceberg</h3>
+<p>Iceberg is a high-performance format for huge analytic tables. Iceberg
brings the reliability and simplicity of SQL tables to big data, while making
it possible for engines like Drill to safely work with the same tables, at the
same time. In addition to being able to query data directly from Iceberg
tables, Drill also allows users to query the Iceberg table metadata as well as
snapshots. <a
href="https://drill.apache.org/docs/iceberg-format-plugin/";>Complete
documentation is availabl [...]
+
+<h3 id="querying-delta-lake">Querying Delta Lake</h3>
+<p>Lest we offend someone, we’re not going to get into the debate between
Iceberg and Delta lake (after all, let’s not argue about who killed whom), but
Delta Lake, if you aren’t familiar with it, is another modern table format
which allows ACID transactions, versioning etc. In version 1.21, Drill adds
support for Delta Lake tables, so users can query Delta Lake tables as well as
associated metadata. You can also query specific versions of files in delta
lake. <a href="https://drill.apa [...]
+
+<h3 id="accessing-access">Accessing Access</h3>
+<p>A surprising number of people use Microsoft Access as a database for their
business data. With version 1.21, Apache Drill can now natively query Microsoft
Access database files using Drill. This can be a major benefit for those
looking to migrate data from Access into more modern formats such as parquet or
even other relational databases. Drill will support Access files from version
1997 and up.</p>
+
+<h3 id="oh-sheets">Oh Sheets!</h3>
+<p>In addition to all of the above, Drill can now query data directly from
GoogleSheets. In addition to being able to query this data source, Drill can
read, write, delete and append to GoogleSheets. Google doesn’t make it easy, so
if this is a feature you are interested in, you’ll definitely want to <a
href="https://drill.apache.org/docs/google-sheets-storage-plugin/";>read the
documentation here</a>.</p>
+
+<h3 id="remote-data">Remote Data</h3>
+<p>As you can see, Drill has significantly expanded the number of data sources
and types that it can query. A part of this work has also been to improve the
implementation behind filesystems. As a result, Drill can now query data stored
on Dropbox, and Box. We added support for filesystems which use OAuth 2.0 for
authorization so this means that more extended file systems are likely coming
your way for the next release.</p>
+
+<h2 id="greatly-improved-access-controls">Greatly Improved Access Controls</h2>
+<p>Managing access controls and credentials on a federated query engine is a
complicated task. Drill has supported a concept called user impersonation which
basically means that Drill can execute queries using the credentials of the
logged in user. This concept works well for querying file systems such as
Hadoop, and other data sources that have the same concepts, however it does not
work at all with data sources that have different concepts of users, or in the
case of OAuth enabled plug [...]
+
+<p>To answer this challenge, Drill 1.21 introduces the concept of user
translation. The idea of user translation is that, when enabled, every user
will have their own unique credentials for specific data sources. Thus, when
that user queries a specific data source, that user’s credentials are used to
execute the query. This is configurable on an individual data source basis.
Ultimately, what this means is that you no longer have to create service
accounts to access data via Drill.</p>
+
+<h2 id="drilling-across-the-clouds">Drilling Across the Clouds</h2>
+<p>While we’re on the subject of clouds, as you may be aware, Drill can query
data stored in cloud-based file systems such as S3, Azure, GCP etc. One of the
challenges however, is that if you have data stored in multiple clouds, it can
become very inefficient to query this data, especially from the perspective of
network IO. As of Drill 1.21, Drill adds a storage plugin which we are calling
Drill on Drill.</p>
+
+<p>Let’s say that you had a Drill cluster in S3, but you had data in both S3
and Azure. With the new Drill on Drill capability, you could install an
additional Drill cluster in Azure, then query both from either Drill cluster.
The advantage is that the queries would be pushed down to the Drill cluster
where the data resides. So if you query Azure from S3, you aren’t sending tons
of data back and forth.</p>
+
+<h2 id="drill-now-supports-more-bi-operators">Drill Now Supports More BI
Operators</h2>
+<p>While Drill held more or less to the SQL standard, it was missing some BI
operators that had become commonplace among SQL platforms. Drill 1.21
introduces the <code class="language-plaintext highlighter-rouge">PIVOT</code>,
and <code class="language-plaintext highlighter-rouge">UNPIVOT</code> operators
which covert rows to columns or vice versa, much in the same way a pivot table
works in Excel. Additionally, we added set operators <code
class="language-plaintext highlighter-rouge">IN [...]
+
+<h2 id="new-statistical-functions">New Statistical Functions</h2>
+<p>Drill 1.21 adds new SQL functions for statistical summaries including <code
class="language-plaintext highlighter-rouge">kendall_correlation</code> for
calculating correlation coefficients, <code class="language-plaintext
highlighter-rouge">width_bucket</code> which is a SQL function for computing
histograms and distributions, and two other functions for computing regression
lines.</p>
+
+<p>Lastly, we’ve also added additional date/time manipulation functions which
will make working with dates significantly easier.</p>
+
+<h2 id="whats-next">What’s Next?</h2>
+<p>The big question is where do we go from here? We’ve already started working
on adding support for additional BI operators such as <code
class="language-plaintext highlighter-rouge">CUBE</code>, <code
class="language-plaintext highlighter-rouge">GROUPING SETS</code> and <code
class="language-plaintext highlighter-rouge">ROLLUP</code>, as well as <code
class="language-plaintext highlighter-rouge">REGEXP_EXTRACT</code>. Since the
new version of Calcite has support for numerous optimizati [...]
diff --git a/output/blog/index.html b/output/blog/index.html
index db8d1f8f9..380583944 100644
--- a/output/blog/index.html
+++ b/output/blog/index.html
@@ -140,6 +140,11 @@
</div>
<div class="int_text" align="left"><!-- previously: site.posts -->
+<p><a class="post-link"
href="/blog/2023/03/02/2023-03-02-drill-1.21-announcement/"></a><br/>
+<span class="post-date">Posted on Mar 2, 2023
+</span>
+<br/></p>
+<!-- previously: site.posts -->
<p><a class="post-link" href="/blog/2023/02/21/drill-1.21.0-released/">Drill
1.21.0 Released</a><br/>
<span class="post-date">Posted on Feb 21, 2023
by James Turton</span>
diff --git a/output/feed.xml b/output/feed.xml
index f90e1812c..ca51552c9 100644
--- a/output/feed.xml
+++ b/output/feed.xml
@@ -6,10 +6,87 @@
</description>
<link>/</link>
<atom:link href="/feed.xml" rel="self" type="application/rss+xml"/>
- <pubDate>Sun, 26 Feb 2023 16:18:34 +0000</pubDate>
- <lastBuildDate>Sun, 26 Feb 2023 16:18:34 +0000</lastBuildDate>
+ <pubDate>Thu, 02 Mar 2023 13:16:17 +0000</pubDate>
+ <lastBuildDate>Thu, 02 Mar 2023 13:16:17 +0000</lastBuildDate>
<generator>Jekyll v3.9.1</generator>
+ <item>
+ <title></title>
+ <description><h1
id="announcing-drill-121-new-connectors-functions-and-much-better-stability">Announcing
Drill 1.21: New Connectors, Functions and Much Better Stability</h1>
+<p>The Apache Drill PMC is pleased to announce a milestone release of
Apache Drill. Since the last release of Drill the team has been hard at work
quashing bugs and making overall functionality improvements. The TL;DR includes
the following:</p>
+
+<ul>
+ <li>New connectors including Apache Iceberg, Delta Lake, Microsoft
Access, GoogleSheets, and Box</li>
+ <li>Efficient cross-cloud query capability</li>
+ <li>Greatly improved access controls to include user translation
support for all storage plugins</li>
+ <li>Greatly improved query planning and implicit casting.</li>
+ <li>New BI-focused SQL operators including <code
class="language-plaintext highlighter-rouge">PIVOT</code>,
<code class="language-plaintext
highlighter-rouge">UNPIVOT</code>, <code
class="language-plaintext highlighter-rouge">EXCEPT</code>
and <code class="language-plaintext
highlighter-rouge">INTERSECT</code></li>
+ <li>New functions for computing regression lines and trends.</li>
+ <li>New and updated date manipulation functions.</li>
+</ul>
+
+<p>Overall, Drill 1.21 is much more capable and stable than previous
versions.</p>
+
+<h2 id="calcite-were-back">Calcite, We’re Back!</h2>
+<p>Drill relies on another open source project, Apache Calcite for its
query planning. The query planning process is a huge part of the overall
functionality of Drill. Unfortunately, about three years ago, there were some
issues in Calcite which forced Drill to fork it and rely on that fork. As a
result, Drill was essentially stuck with a three year old query planner, but
more importantly, bugs that were fixed in Calcite, as well as new capabilities
were not finding their way into [...]
+
+<p>That is no longer the case. Drill 1.21 is now running on the latest
stable version of Calcite, version 1.33. As a result, we’ve been able to close
countless JIRA tickets of various queries failing and other random bugs that
were the result of query planning bugs.</p>
+
+<p>What this means for you as a user is that you’ll see much fewer
queries failing and better overall performance in terms of speed and stability.
You’ll see better optimizations being pushed down to JDBC data sources as well
as support for BigQuery, Athena and other JDBC data sources. We hope to keep
Drill away from Calcite forks so I hope that we will work with the Calcite
community to keep our tools in sync.</p>
+
+<h2
id="improved-implicit-casting-rules-reduce-schema-change-failures">Improved
Implicit Casting Rules Reduce Schema Change Failures</h2>
+<p>From this author’s perspective, one of the biggest improvements in
Drill is one of the least noticeable and that is the result of improved
implicit casting. One of Drill’s unique features is its ability to infer the
structure, or schema of your data. However, this can be problematic when the
schema changes. When I used to teach Drill, I used to have spend a considerable
amount of time teaching students how to cast data from one data type to another
to ensure that the queries wou [...]
+
+<p>When using latest version of Drill, you’ll find that queries will
work without the need for much if any casting. In short, they’ll do what you
expect them to do. It’s really a high on magic functionality.</p>
+
+<h2
id="integrations-with-the-modern-and-not-so-modern-data-stack">Integrations
with the Modern and Not-so-Modern Data Stack</h2>
+<p>The new version of Drill features several new connectors and readers
that will enable users to connect to the “modern data stack”, specifically
support for Apache Iceberg and Delta Lake.</p>
+
+<h3 id="breaking-the-iceberg">Breaking the Iceberg</h3>
+<p>Iceberg is a high-performance format for huge analytic tables.
Iceberg brings the reliability and simplicity of SQL tables to big data, while
making it possible for engines like Drill to safely work with the same tables,
at the same time. In addition to being able to query data directly from Iceberg
tables, Drill also allows users to query the Iceberg table metadata as well as
snapshots. <a
href="https://drill.apache.org/docs/iceberg-format-plugin/">Complete
doc [...]
+
+<h3 id="querying-delta-lake">Querying Delta Lake</h3>
+<p>Lest we offend someone, we’re not going to get into the debate
between Iceberg and Delta lake (after all, let’s not argue about who killed
whom), but Delta Lake, if you aren’t familiar with it, is another modern table
format which allows ACID transactions, versioning etc. In version 1.21, Drill
adds support for Delta Lake tables, so users can query Delta Lake tables as
well as associated metadata. You can also query specific versions of files in
delta lake. <a href="htt [...]
+
+<h3 id="accessing-access">Accessing Access</h3>
+<p>A surprising number of people use Microsoft Access as a database for
their business data. With version 1.21, Apache Drill can now natively query
Microsoft Access database files using Drill. This can be a major benefit for
those looking to migrate data from Access into more modern formats such as
parquet or even other relational databases. Drill will support Access files
from version 1997 and up.</p>
+
+<h3 id="oh-sheets">Oh Sheets!</h3>
+<p>In addition to all of the above, Drill can now query data directly
from GoogleSheets. In addition to being able to query this data source, Drill
can read, write, delete and append to GoogleSheets. Google doesn’t make it
easy, so if this is a feature you are interested in, you’ll definitely want to
<a
href="https://drill.apache.org/docs/google-sheets-storage-plugin/">read
the documentation here</a>.</p>
+
+<h3 id="remote-data">Remote Data</h3>
+<p>As you can see, Drill has significantly expanded the number of data
sources and types that it can query. A part of this work has also been to
improve the implementation behind filesystems. As a result, Drill can now query
data stored on Dropbox, and Box. We added support for filesystems which use
OAuth 2.0 for authorization so this means that more extended file systems are
likely coming your way for the next release.</p>
+
+<h2 id="greatly-improved-access-controls">Greatly Improved
Access Controls</h2>
+<p>Managing access controls and credentials on a federated query engine
is a complicated task. Drill has supported a concept called user impersonation
which basically means that Drill can execute queries using the credentials of
the logged in user. This concept works well for querying file systems such as
Hadoop, and other data sources that have the same concepts, however it does not
work at all with data sources that have different concepts of users, or in the
case of OAuth enable [...]
+
+<p>To answer this challenge, Drill 1.21 introduces the concept of user
translation. The idea of user translation is that, when enabled, every user
will have their own unique credentials for specific data sources. Thus, when
that user queries a specific data source, that user’s credentials are used to
execute the query. This is configurable on an individual data source basis.
Ultimately, what this means is that you no longer have to create service
accounts to access data via Drill.& [...]
+
+<h2 id="drilling-across-the-clouds">Drilling Across the
Clouds</h2>
+<p>While we’re on the subject of clouds, as you may be aware, Drill can
query data stored in cloud-based file systems such as S3, Azure, GCP etc. One
of the challenges however, is that if you have data stored in multiple clouds,
it can become very inefficient to query this data, especially from the
perspective of network IO. As of Drill 1.21, Drill adds a storage plugin which
we are calling Drill on Drill.</p>
+
+<p>Let’s say that you had a Drill cluster in S3, but you had data in
both S3 and Azure. With the new Drill on Drill capability, you could install an
additional Drill cluster in Azure, then query both from either Drill cluster.
The advantage is that the queries would be pushed down to the Drill cluster
where the data resides. So if you query Azure from S3, you aren’t sending tons
of data back and forth.</p>
+
+<h2 id="drill-now-supports-more-bi-operators">Drill Now
Supports More BI Operators</h2>
+<p>While Drill held more or less to the SQL standard, it was missing
some BI operators that had become commonplace among SQL platforms. Drill 1.21
introduces the <code class="language-plaintext
highlighter-rouge">PIVOT</code>, and <code
class="language-plaintext highlighter-rouge">UNPIVOT</code>
operators which covert rows to columns or vice versa, much in the same way a
pivot table works in Excel. Additionally, we added set operators < [...]
+
+<h2 id="new-statistical-functions">New Statistical
Functions</h2>
+<p>Drill 1.21 adds new SQL functions for statistical summaries including
<code class="language-plaintext
highlighter-rouge">kendall_correlation</code> for calculating
correlation coefficients, <code class="language-plaintext
highlighter-rouge">width_bucket</code> which is a SQL function
for computing histograms and distributions, and two other functions for
computing regression lines.</p>
+
+<p>Lastly, we’ve also added additional date/time manipulation functions
which will make working with dates significantly easier.</p>
+
+<h2 id="whats-next">What’s Next?</h2>
+<p>The big question is where do we go from here? We’ve already started
working on adding support for additional BI operators such as <code
class="language-plaintext highlighter-rouge">CUBE</code>,
<code class="language-plaintext highlighter-rouge">GROUPING
SETS</code> and <code class="language-plaintext
highlighter-rouge">ROLLUP</code>, as well as <code
class="language-plaintext highlighter-rouge">REG [...]
+</description>
+ <pubDate>Thu, 02 Mar 2023 13:16:17 +0000</pubDate>
+ <link>/blog/2023/03/02/2023-03-02-drill-1.21-announcement/</link>
+ <guid
isPermaLink="true">/blog/2023/03/02/2023-03-02-drill-1.21-announcement/</guid>
+
+
+ <category>blog</category>
+
+ </item>
+
<item>
<title>Drill 1.21.0 Released</title>
<description><p>Today, we’re happy to announce the availability
of Drill 1.21.0. You can download it <a
href="https://drill.apache.org/download/">here</a>.</p>
@@ -235,44 +312,6 @@ In [7]: while True:
<guid
isPermaLink="true">/blog/2021/07/09/streaming-data-from-the-rest-api/</guid>
- <category>blog</category>
-
- </item>
-
- <item>
- <title>Drill 1.19 Released</title>
- <description><p>Today, we’re happy to announce the availability
of Drill 1.19.0. You can download it <a
href="https://drill.apache.org/download/">here</a>.</p>
-
-<h2 id="this-release-provides-the-following-new-features">This
release provides the following new Features:</h2>
-
-<ul>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-92">DRILL-92</a>
- Cassandra Storage Plugin</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-3637">DRILL-3637</a>
- Elasticsearch Storage Plugin</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7823">DRILL-7823</a>
- XML Storage Plugin</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7751">DRILL-7751</a>
- Splunk Storage Plugin</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-5940">DRILL-5940</a>
- Avro with schema registry support for Kafka</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7855">DRILL-7855</a>
- Secure mechanism for specifying storage plugin credentials</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7921">DRILL-7921</a>
- Linux ARM64 based system support</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-6953">DRILL-6953</a>
- Rowset based JSON reader</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7733">DRILL-7733</a>
- Use streaming for REST JSON queries</li>
- <li>Several plugins have been converted to the Enhanced Vector
Framework (EVF)
- <ul>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7525">DRILL-7525</a>
- Convert SequenceFiles to EVF</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7532">DRILL-7532</a>
- Convert SysLog to EVF</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7533">DRILL-7533</a>
- Convert Pcapng to EVF</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7534">DRILL-7534</a>
- Convert HTTPD format plugin to EVF</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7536">DRILL-7533</a>
- Convert Image Format to EVF</li>
- </ul>
- </li>
-</ul>
-
-<p>You can find a complete list of improvements and JIRAs resolved in
the 1.19.0 release <a
href="/docs/apache-drill-1-19-0-release-notes/">here</a>.</p>
-</description>
- <pubDate>Thu, 10 Jun 2021 00:00:00 +0000</pubDate>
- <link>/blog/2021/06/10/drill-1.19-released/</link>
- <guid isPermaLink="true">/blog/2021/06/10/drill-1.19-released/</guid>
-
-
<category>blog</category>
</item>
diff --git a/output/index.html b/output/index.html
index 77ee47e74..ebf7e8920 100644
--- a/output/index.html
+++ b/output/index.html
@@ -203,9 +203,9 @@ $(document).ready(function() {
<div class="news">News:
</div>
- <div><a href="/blog/2023/02/21/drill-1.21.0-released/">Drill 1.21.0
Released</a><br/><span>(James Turton)</span></div>
+ <div><a
href="/blog/2023/03/02/2023-03-02-drill-1.21-announcement/"></a><br/><span>()</span></div>
- <div><a href="/blog/2023/01/07/drill-1.20.3-released/">Drill 1.20.3
Released</a><br/><span>(James Turton)</span></div>
+ <div><a href="/blog/2023/02/21/drill-1.21.0-released/">Drill 1.21.0
Released</a><br/><span>(James Turton)</span></div>
</div>
<div class="mw introWrapper">
<table class="intro" cellpadding="0" cellspacing="0" align="center">
diff --git
a/output/zh/blog/2023/03/02/2023-03-02-drill-1.21-announcement/index.html
b/output/zh/blog/2023/03/02/2023-03-02-drill-1.21-announcement/index.html
new file mode 100644
index 000000000..6a3135e20
--- /dev/null
+++ b/output/zh/blog/2023/03/02/2023-03-02-drill-1.21-announcement/index.html
@@ -0,0 +1,65 @@
+<h1
id="announcing-drill-121-new-connectors-functions-and-much-better-stability">Announcing
Drill 1.21: New Connectors, Functions and Much Better Stability</h1>
+<p>The Apache Drill PMC is pleased to announce a milestone release of Apache
Drill. Since the last release of Drill the team has been hard at work quashing
bugs and making overall functionality improvements. The TL;DR includes the
following:</p>
+
+<ul>
+ <li>New connectors including Apache Iceberg, Delta Lake, Microsoft Access,
GoogleSheets, and Box</li>
+ <li>Efficient cross-cloud query capability</li>
+ <li>Greatly improved access controls to include user translation support for
all storage plugins</li>
+ <li>Greatly improved query planning and implicit casting.</li>
+ <li>New BI-focused SQL operators including <code class="language-plaintext
highlighter-rouge">PIVOT</code>, <code class="language-plaintext
highlighter-rouge">UNPIVOT</code>, <code class="language-plaintext
highlighter-rouge">EXCEPT</code> and <code class="language-plaintext
highlighter-rouge">INTERSECT</code></li>
+ <li>New functions for computing regression lines and trends.</li>
+ <li>New and updated date manipulation functions.</li>
+</ul>
+
+<p>Overall, Drill 1.21 is much more capable and stable than previous
versions.</p>
+
+<h2 id="calcite-were-back">Calcite, We’re Back!</h2>
+<p>Drill relies on another open source project, Apache Calcite for its query
planning. The query planning process is a huge part of the overall
functionality of Drill. Unfortunately, about three years ago, there were some
issues in Calcite which forced Drill to fork it and rely on that fork. As a
result, Drill was essentially stuck with a three year old query planner, but
more importantly, bugs that were fixed in Calcite, as well as new capabilities
were not finding their way into Drill.</p>
+
+<p>That is no longer the case. Drill 1.21 is now running on the latest stable
version of Calcite, version 1.33. As a result, we’ve been able to close
countless JIRA tickets of various queries failing and other random bugs that
were the result of query planning bugs.</p>
+
+<p>What this means for you as a user is that you’ll see much fewer queries
failing and better overall performance in terms of speed and stability. You’ll
see better optimizations being pushed down to JDBC data sources as well as
support for BigQuery, Athena and other JDBC data sources. We hope to keep Drill
away from Calcite forks so I hope that we will work with the Calcite community
to keep our tools in sync.</p>
+
+<h2
id="improved-implicit-casting-rules-reduce-schema-change-failures">Improved
Implicit Casting Rules Reduce Schema Change Failures</h2>
+<p>From this author’s perspective, one of the biggest improvements in Drill is
one of the least noticeable and that is the result of improved implicit
casting. One of Drill’s unique features is its ability to infer the structure,
or schema of your data. However, this can be problematic when the schema
changes. When I used to teach Drill, I used to have spend a considerable amount
of time teaching students how to cast data from one data type to another to
ensure that the queries would suc [...]
+
+<p>When using latest version of Drill, you’ll find that queries will work
without the need for much if any casting. In short, they’ll do what you expect
them to do. It’s really a high on magic functionality.</p>
+
+<h2
id="integrations-with-the-modern-and-not-so-modern-data-stack">Integrations
with the Modern and Not-so-Modern Data Stack</h2>
+<p>The new version of Drill features several new connectors and readers that
will enable users to connect to the “modern data stack”, specifically support
for Apache Iceberg and Delta Lake.</p>
+
+<h3 id="breaking-the-iceberg">Breaking the Iceberg</h3>
+<p>Iceberg is a high-performance format for huge analytic tables. Iceberg
brings the reliability and simplicity of SQL tables to big data, while making
it possible for engines like Drill to safely work with the same tables, at the
same time. In addition to being able to query data directly from Iceberg
tables, Drill also allows users to query the Iceberg table metadata as well as
snapshots. <a
href="https://drill.apache.org/docs/iceberg-format-plugin/";>Complete
documentation is availabl [...]
+
+<h3 id="querying-delta-lake">Querying Delta Lake</h3>
+<p>Lest we offend someone, we’re not going to get into the debate between
Iceberg and Delta lake (after all, let’s not argue about who killed whom), but
Delta Lake, if you aren’t familiar with it, is another modern table format
which allows ACID transactions, versioning etc. In version 1.21, Drill adds
support for Delta Lake tables, so users can query Delta Lake tables as well as
associated metadata. You can also query specific versions of files in delta
lake. <a href="https://drill.apa [...]
+
+<h3 id="accessing-access">Accessing Access</h3>
+<p>A surprising number of people use Microsoft Access as a database for their
business data. With version 1.21, Apache Drill can now natively query Microsoft
Access database files using Drill. This can be a major benefit for those
looking to migrate data from Access into more modern formats such as parquet or
even other relational databases. Drill will support Access files from version
1997 and up.</p>
+
+<h3 id="oh-sheets">Oh Sheets!</h3>
+<p>In addition to all of the above, Drill can now query data directly from
GoogleSheets. In addition to being able to query this data source, Drill can
read, write, delete and append to GoogleSheets. Google doesn’t make it easy, so
if this is a feature you are interested in, you’ll definitely want to <a
href="https://drill.apache.org/docs/google-sheets-storage-plugin/";>read the
documentation here</a>.</p>
+
+<h3 id="remote-data">Remote Data</h3>
+<p>As you can see, Drill has significantly expanded the number of data sources
and types that it can query. A part of this work has also been to improve the
implementation behind filesystems. As a result, Drill can now query data stored
on Dropbox, and Box. We added support for filesystems which use OAuth 2.0 for
authorization so this means that more extended file systems are likely coming
your way for the next release.</p>
+
+<h2 id="greatly-improved-access-controls">Greatly Improved Access Controls</h2>
+<p>Managing access controls and credentials on a federated query engine is a
complicated task. Drill has supported a concept called user impersonation which
basically means that Drill can execute queries using the credentials of the
logged in user. This concept works well for querying file systems such as
Hadoop, and other data sources that have the same concepts, however it does not
work at all with data sources that have different concepts of users, or in the
case of OAuth enabled plug [...]
+
+<p>To answer this challenge, Drill 1.21 introduces the concept of user
translation. The idea of user translation is that, when enabled, every user
will have their own unique credentials for specific data sources. Thus, when
that user queries a specific data source, that user’s credentials are used to
execute the query. This is configurable on an individual data source basis.
Ultimately, what this means is that you no longer have to create service
accounts to access data via Drill.</p>
+
+<h2 id="drilling-across-the-clouds">Drilling Across the Clouds</h2>
+<p>While we’re on the subject of clouds, as you may be aware, Drill can query
data stored in cloud-based file systems such as S3, Azure, GCP etc. One of the
challenges however, is that if you have data stored in multiple clouds, it can
become very inefficient to query this data, especially from the perspective of
network IO. As of Drill 1.21, Drill adds a storage plugin which we are calling
Drill on Drill.</p>
+
+<p>Let’s say that you had a Drill cluster in S3, but you had data in both S3
and Azure. With the new Drill on Drill capability, you could install an
additional Drill cluster in Azure, then query both from either Drill cluster.
The advantage is that the queries would be pushed down to the Drill cluster
where the data resides. So if you query Azure from S3, you aren’t sending tons
of data back and forth.</p>
+
+<h2 id="drill-now-supports-more-bi-operators">Drill Now Supports More BI
Operators</h2>
+<p>While Drill held more or less to the SQL standard, it was missing some BI
operators that had become commonplace among SQL platforms. Drill 1.21
introduces the <code class="language-plaintext highlighter-rouge">PIVOT</code>,
and <code class="language-plaintext highlighter-rouge">UNPIVOT</code> operators
which covert rows to columns or vice versa, much in the same way a pivot table
works in Excel. Additionally, we added set operators <code
class="language-plaintext highlighter-rouge">IN [...]
+
+<h2 id="new-statistical-functions">New Statistical Functions</h2>
+<p>Drill 1.21 adds new SQL functions for statistical summaries including <code
class="language-plaintext highlighter-rouge">kendall_correlation</code> for
calculating correlation coefficients, <code class="language-plaintext
highlighter-rouge">width_bucket</code> which is a SQL function for computing
histograms and distributions, and two other functions for computing regression
lines.</p>
+
+<p>Lastly, we’ve also added additional date/time manipulation functions which
will make working with dates significantly easier.</p>
+
+<h2 id="whats-next">What’s Next?</h2>
+<p>The big question is where do we go from here? We’ve already started working
on adding support for additional BI operators such as <code
class="language-plaintext highlighter-rouge">CUBE</code>, <code
class="language-plaintext highlighter-rouge">GROUPING SETS</code> and <code
class="language-plaintext highlighter-rouge">ROLLUP</code>, as well as <code
class="language-plaintext highlighter-rouge">REGEXP_EXTRACT</code>. Since the
new version of Calcite has support for numerous optimizati [...]
diff --git a/output/zh/blog/index.html b/output/zh/blog/index.html
index cb8e6d263..2e0c3d5b9 100644
--- a/output/zh/blog/index.html
+++ b/output/zh/blog/index.html
@@ -140,6 +140,11 @@
</div>
<div class="int_text" align="left"><!-- previously: site.posts -->
+<p><a class="post-link"
href="/zh/blog/2023/03/02/2023-03-02-drill-1.21-announcement/"></a><br/>
+<span class="post-date">Posted on Mar 2, 2023
+</span>
+<br/></p>
+<!-- previously: site.posts -->
<p><a class="post-link"
href="/zh/blog/2023/02/21/drill-1.21.0-released/">Drill 1.21.0 Released</a><br/>
<span class="post-date">Posted on Feb 21, 2023
by James Turton</span>
diff --git a/output/zh/feed.xml b/output/zh/feed.xml
index 0236010ed..2ac4925a6 100644
--- a/output/zh/feed.xml
+++ b/output/zh/feed.xml
@@ -6,10 +6,87 @@
</description>
<link>/</link>
<atom:link href="/zh/feed.xml" rel="self" type="application/rss+xml"/>
- <pubDate>Sun, 26 Feb 2023 16:18:34 +0000</pubDate>
- <lastBuildDate>Sun, 26 Feb 2023 16:18:34 +0000</lastBuildDate>
+ <pubDate>Thu, 02 Mar 2023 13:16:17 +0000</pubDate>
+ <lastBuildDate>Thu, 02 Mar 2023 13:16:17 +0000</lastBuildDate>
<generator>Jekyll v3.9.1</generator>
+ <item>
+ <title></title>
+ <description><h1
id="announcing-drill-121-new-connectors-functions-and-much-better-stability">Announcing
Drill 1.21: New Connectors, Functions and Much Better Stability</h1>
+<p>The Apache Drill PMC is pleased to announce a milestone release of
Apache Drill. Since the last release of Drill the team has been hard at work
quashing bugs and making overall functionality improvements. The TL;DR includes
the following:</p>
+
+<ul>
+ <li>New connectors including Apache Iceberg, Delta Lake, Microsoft
Access, GoogleSheets, and Box</li>
+ <li>Efficient cross-cloud query capability</li>
+ <li>Greatly improved access controls to include user translation
support for all storage plugins</li>
+ <li>Greatly improved query planning and implicit casting.</li>
+ <li>New BI-focused SQL operators including <code
class="language-plaintext highlighter-rouge">PIVOT</code>,
<code class="language-plaintext
highlighter-rouge">UNPIVOT</code>, <code
class="language-plaintext highlighter-rouge">EXCEPT</code>
and <code class="language-plaintext
highlighter-rouge">INTERSECT</code></li>
+ <li>New functions for computing regression lines and trends.</li>
+ <li>New and updated date manipulation functions.</li>
+</ul>
+
+<p>Overall, Drill 1.21 is much more capable and stable than previous
versions.</p>
+
+<h2 id="calcite-were-back">Calcite, We’re Back!</h2>
+<p>Drill relies on another open source project, Apache Calcite for its
query planning. The query planning process is a huge part of the overall
functionality of Drill. Unfortunately, about three years ago, there were some
issues in Calcite which forced Drill to fork it and rely on that fork. As a
result, Drill was essentially stuck with a three year old query planner, but
more importantly, bugs that were fixed in Calcite, as well as new capabilities
were not finding their way into [...]
+
+<p>That is no longer the case. Drill 1.21 is now running on the latest
stable version of Calcite, version 1.33. As a result, we’ve been able to close
countless JIRA tickets of various queries failing and other random bugs that
were the result of query planning bugs.</p>
+
+<p>What this means for you as a user is that you’ll see much fewer
queries failing and better overall performance in terms of speed and stability.
You’ll see better optimizations being pushed down to JDBC data sources as well
as support for BigQuery, Athena and other JDBC data sources. We hope to keep
Drill away from Calcite forks so I hope that we will work with the Calcite
community to keep our tools in sync.</p>
+
+<h2
id="improved-implicit-casting-rules-reduce-schema-change-failures">Improved
Implicit Casting Rules Reduce Schema Change Failures</h2>
+<p>From this author’s perspective, one of the biggest improvements in
Drill is one of the least noticeable and that is the result of improved
implicit casting. One of Drill’s unique features is its ability to infer the
structure, or schema of your data. However, this can be problematic when the
schema changes. When I used to teach Drill, I used to have spend a considerable
amount of time teaching students how to cast data from one data type to another
to ensure that the queries wou [...]
+
+<p>When using latest version of Drill, you’ll find that queries will
work without the need for much if any casting. In short, they’ll do what you
expect them to do. It’s really a high on magic functionality.</p>
+
+<h2
id="integrations-with-the-modern-and-not-so-modern-data-stack">Integrations
with the Modern and Not-so-Modern Data Stack</h2>
+<p>The new version of Drill features several new connectors and readers
that will enable users to connect to the “modern data stack”, specifically
support for Apache Iceberg and Delta Lake.</p>
+
+<h3 id="breaking-the-iceberg">Breaking the Iceberg</h3>
+<p>Iceberg is a high-performance format for huge analytic tables.
Iceberg brings the reliability and simplicity of SQL tables to big data, while
making it possible for engines like Drill to safely work with the same tables,
at the same time. In addition to being able to query data directly from Iceberg
tables, Drill also allows users to query the Iceberg table metadata as well as
snapshots. <a
href="https://drill.apache.org/docs/iceberg-format-plugin/">Complete
doc [...]
+
+<h3 id="querying-delta-lake">Querying Delta Lake</h3>
+<p>Lest we offend someone, we’re not going to get into the debate
between Iceberg and Delta lake (after all, let’s not argue about who killed
whom), but Delta Lake, if you aren’t familiar with it, is another modern table
format which allows ACID transactions, versioning etc. In version 1.21, Drill
adds support for Delta Lake tables, so users can query Delta Lake tables as
well as associated metadata. You can also query specific versions of files in
delta lake. <a href="htt [...]
+
+<h3 id="accessing-access">Accessing Access</h3>
+<p>A surprising number of people use Microsoft Access as a database for
their business data. With version 1.21, Apache Drill can now natively query
Microsoft Access database files using Drill. This can be a major benefit for
those looking to migrate data from Access into more modern formats such as
parquet or even other relational databases. Drill will support Access files
from version 1997 and up.</p>
+
+<h3 id="oh-sheets">Oh Sheets!</h3>
+<p>In addition to all of the above, Drill can now query data directly
from GoogleSheets. In addition to being able to query this data source, Drill
can read, write, delete and append to GoogleSheets. Google doesn’t make it
easy, so if this is a feature you are interested in, you’ll definitely want to
<a
href="https://drill.apache.org/docs/google-sheets-storage-plugin/">read
the documentation here</a>.</p>
+
+<h3 id="remote-data">Remote Data</h3>
+<p>As you can see, Drill has significantly expanded the number of data
sources and types that it can query. A part of this work has also been to
improve the implementation behind filesystems. As a result, Drill can now query
data stored on Dropbox, and Box. We added support for filesystems which use
OAuth 2.0 for authorization so this means that more extended file systems are
likely coming your way for the next release.</p>
+
+<h2 id="greatly-improved-access-controls">Greatly Improved
Access Controls</h2>
+<p>Managing access controls and credentials on a federated query engine
is a complicated task. Drill has supported a concept called user impersonation
which basically means that Drill can execute queries using the credentials of
the logged in user. This concept works well for querying file systems such as
Hadoop, and other data sources that have the same concepts, however it does not
work at all with data sources that have different concepts of users, or in the
case of OAuth enable [...]
+
+<p>To answer this challenge, Drill 1.21 introduces the concept of user
translation. The idea of user translation is that, when enabled, every user
will have their own unique credentials for specific data sources. Thus, when
that user queries a specific data source, that user’s credentials are used to
execute the query. This is configurable on an individual data source basis.
Ultimately, what this means is that you no longer have to create service
accounts to access data via Drill.& [...]
+
+<h2 id="drilling-across-the-clouds">Drilling Across the
Clouds</h2>
+<p>While we’re on the subject of clouds, as you may be aware, Drill can
query data stored in cloud-based file systems such as S3, Azure, GCP etc. One
of the challenges however, is that if you have data stored in multiple clouds,
it can become very inefficient to query this data, especially from the
perspective of network IO. As of Drill 1.21, Drill adds a storage plugin which
we are calling Drill on Drill.</p>
+
+<p>Let’s say that you had a Drill cluster in S3, but you had data in
both S3 and Azure. With the new Drill on Drill capability, you could install an
additional Drill cluster in Azure, then query both from either Drill cluster.
The advantage is that the queries would be pushed down to the Drill cluster
where the data resides. So if you query Azure from S3, you aren’t sending tons
of data back and forth.</p>
+
+<h2 id="drill-now-supports-more-bi-operators">Drill Now
Supports More BI Operators</h2>
+<p>While Drill held more or less to the SQL standard, it was missing
some BI operators that had become commonplace among SQL platforms. Drill 1.21
introduces the <code class="language-plaintext
highlighter-rouge">PIVOT</code>, and <code
class="language-plaintext highlighter-rouge">UNPIVOT</code>
operators which covert rows to columns or vice versa, much in the same way a
pivot table works in Excel. Additionally, we added set operators < [...]
+
+<h2 id="new-statistical-functions">New Statistical
Functions</h2>
+<p>Drill 1.21 adds new SQL functions for statistical summaries including
<code class="language-plaintext
highlighter-rouge">kendall_correlation</code> for calculating
correlation coefficients, <code class="language-plaintext
highlighter-rouge">width_bucket</code> which is a SQL function
for computing histograms and distributions, and two other functions for
computing regression lines.</p>
+
+<p>Lastly, we’ve also added additional date/time manipulation functions
which will make working with dates significantly easier.</p>
+
+<h2 id="whats-next">What’s Next?</h2>
+<p>The big question is where do we go from here? We’ve already started
working on adding support for additional BI operators such as <code
class="language-plaintext highlighter-rouge">CUBE</code>,
<code class="language-plaintext highlighter-rouge">GROUPING
SETS</code> and <code class="language-plaintext
highlighter-rouge">ROLLUP</code>, as well as <code
class="language-plaintext highlighter-rouge">REG [...]
+</description>
+ <pubDate>Thu, 02 Mar 2023 13:16:17 +0000</pubDate>
+ <link>/blog/2023/03/02/2023-03-02-drill-1.21-announcement/</link>
+ <guid
isPermaLink="true">/blog/2023/03/02/2023-03-02-drill-1.21-announcement/</guid>
+
+
+ <category>blog</category>
+
+ </item>
+
<item>
<title>Drill 1.21.0 Released</title>
<description><p>Today, we’re happy to announce the availability
of Drill 1.21.0. You can download it <a
href="https://drill.apache.org/download/">here</a>.</p>
@@ -235,44 +312,6 @@ In [7]: while True:
<guid
isPermaLink="true">/blog/2021/07/09/streaming-data-from-the-rest-api/</guid>
- <category>blog</category>
-
- </item>
-
- <item>
- <title>Drill 1.19 Released</title>
- <description><p>今天, we’re happy to announce the availability of
Drill 1.19.0. You can download it <a
href="https://drill.apache.org/download/">here</a>.</p>
-
-<h2 id="this-release-provides-the-following-new-features">This
release provides the following new Features:</h2>
-
-<ul>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-92">DRILL-92</a>
- Cassandra Storage Plugin</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-3637">DRILL-3637</a>
- Elasticsearch Storage Plugin</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7823">DRILL-7823</a>
- XML Storage Plugin</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7751">DRILL-7751</a>
- Splunk Storage Plugin</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-5940">DRILL-5940</a>
- Avro with schema registry support for Kafka</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7855">DRILL-7855</a>
- Secure mechanism for specifying storage plugin credentials</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7921">DRILL-7921</a>
- Linux ARM64 based system support</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-6953">DRILL-6953</a>
- Rowset based JSON reader</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7733">DRILL-7733</a>
- Use streaming for REST JSON queries</li>
- <li>Several plugins have been converted to the Enhanced Vector
Framework (EVF)
- <ul>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7525">DRILL-7525</a>
- Convert SequenceFiles to EVF</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7532">DRILL-7532</a>
- Convert SysLog to EVF</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7533">DRILL-7533</a>
- Convert Pcapng to EVF</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7534">DRILL-7534</a>
- Convert HTTPD format plugin to EVF</li>
- <li><a
href="https://issues.apache.org/jira/browse/DRILL-7536">DRILL-7533</a>
- Convert Image Format to EVF</li>
- </ul>
- </li>
-</ul>
-
-<p>You can find a complete list of improvements and JIRAs resolved in
the 1.19.0 release <a
href="/docs/apache-drill-1-19-0-release-notes/">here</a>.</p>
-</description>
- <pubDate>Thu, 10 Jun 2021 00:00:00 +0000</pubDate>
- <link>/blog/2021/06/10/drill-1.19-released/</link>
- <guid isPermaLink="true">/blog/2021/06/10/drill-1.19-released/</guid>
-
-
<category>blog</category>
</item>
diff --git a/output/zh/index.html b/output/zh/index.html
index 985b164e6..3be82f251 100644
--- a/output/zh/index.html
+++ b/output/zh/index.html
@@ -202,9 +202,9 @@
<div class="news">News:
</div>
- <div><a href="/zh/blog/2023/02/21/drill-1.21.0-released/">Drill 1.21.0
Released</a><br/><span>(James Turton)</span></div>
+ <div><a
href="/zh/blog/2023/03/02/2023-03-02-drill-1.21-announcement/"></a><br/><span>()</span></div>
- <div><a href="/zh/blog/2023/01/07/drill-1.20.3-released/">Drill 1.20.3
Released</a><br/><span>(James Turton)</span></div>
+ <div><a href="/zh/blog/2023/02/21/drill-1.21.0-released/">Drill 1.21.0
Released</a><br/><span>(James Turton)</span></div>
</div>
<div class="mw introWrapper">
<table class="intro" cellpadding="0" cellspacing="0" align="center">
