<?xml version="1.0" encoding="utf-8"?>
<feed xml:lang="en" xmlns="http://www.w3.org/2005/Atom"><title>Recent changes to patches</title><link href="https://sourceforge.net/p/scribeserver/patches/" rel="alternate"/><link href="https://sourceforge.net/p/scribeserver/patches/feed.atom" rel="self"/><id>https://sourceforge.net/p/scribeserver/patches/</id><updated>2009-08-21T17:15:16Z</updated><subtitle>Recent changes to patches</subtitle><entry><title>Added minute-granularity to time-based log rotate for stores</title><link href="https://sourceforge.net/p/scribeserver/patches/11/" rel="alternate"/><published>2009-08-21T17:15:16Z</published><updated>2009-08-21T17:15:16Z</updated><author><name>Elbert</name><uri>https://sourceforge.net/u/mudder73/</uri></author><id>https://sourceforge.net5b4de970c9d0704b02a4917d534245e0c5f4e94e</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;The stable distribution allows time-based rotation values to be never, daily, and hourly. I have need of them to be rotated every X minutes as well, where X &amp;lt; 60. To enable this, in the scribe config file, set rotate_period=minutely and then set rotate_minute_interval to some value. If rotate_minute_interval is omitted, or is not &amp;gt;0 &amp;amp;&amp;amp; &amp;lt;60, it will default to 15, meaning the log file will rotate every 15 minutes at 0, 15, 30, and 45.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>scribe svn rev 11 fails to compile</title><link href="https://sourceforge.net/p/scribeserver/patches/10/" rel="alternate"/><published>2009-08-17T06:54:24Z</published><updated>2009-08-17T06:54:24Z</updated><author><name>Jon Schutz</name><uri>https://sourceforge.net/u/jonschutz/</uri></author><id>https://sourceforge.net5c6bb582dab75f0e6d97f644d39dc9735f51a074</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Tried with gcc-4.2.4 and gcc-4.3.2, both fail with error&lt;/p&gt;
&lt;p&gt;store.cpp: In member function ‘long unsigned int BucketStore::bucketize(const std::string&amp;amp;)’:&lt;br /&gt;
store.cpp:2077: error: jump to case label&lt;/p&gt;
&lt;p&gt;Fix is trivial, see svn diff attached.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Daemon mode, SysV startup scripts</title><link href="https://sourceforge.net/p/scribeserver/patches/9/" rel="alternate"/><published>2009-05-08T06:00:58Z</published><updated>2009-05-08T06:00:58Z</updated><author><name>Jon Schutz</name><uri>https://sourceforge.net/u/jonschutz/</uri></author><id>https://sourceforge.net30ac5a97bb5be0cb9864d81ce807d323d4454f6a</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;The attached patch adds a '-d' (daemonise) option to scribed so that it can more easily be invoked as a system service, and also provides SysV startup scripts for RedHat (CentOS/Fedora) and Ubuntu/Debian in the contrib directory.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>improve categories support</title><link href="https://sourceforge.net/p/scribeserver/patches/8/" rel="alternate"/><published>2009-05-04T08:42:36Z</published><updated>2009-05-04T08:42:36Z</updated><author><name>Anthony Giardullo</name><uri>https://sourceforge.net/u/agiardullo/</uri></author><id>https://sourceforge.netf29cd27ed930c23ae3bf48bb10ff8b65eaf77ef4</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Clean up the handling of the new 'categories=' config option.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>add ignored message count</title><link href="https://sourceforge.net/p/scribeserver/patches/7/" rel="alternate"/><published>2009-05-04T08:40:16Z</published><updated>2009-05-04T08:40:16Z</updated><author><name>Anthony Giardullo</name><uri>https://sourceforge.net/u/agiardullo/</uri></author><id>https://sourceforge.net598a5438c72ff0506a486beb1c8e143c7727bf1f</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Adds a counter 'ignored' for Null stores that tracks how many messages have been handled by a null store.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>add replay_buffer option for buffer stores</title><link href="https://sourceforge.net/p/scribeserver/patches/6/" rel="alternate"/><published>2009-05-04T08:39:15Z</published><updated>2009-05-04T08:39:15Z</updated><author><name>Anthony Giardullo</name><uri>https://sourceforge.net/u/agiardullo/</uri></author><id>https://sourceforge.net1f1794bb8e5d1d7261dacfff490d17d9580d131a</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;A buffer store will log messages to the secondary store if the primary store is not available.  By default, the buffer store will try to move the messages from the secondary store to the primary store once the primary store is available again.  With 'replay_buffer=no', then scribe will just leave these messages in the secondary store and not try to move them back to the primary store.  (The default behavior is not changed).&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>add key_range hasing to bucket store</title><link href="https://sourceforge.net/p/scribeserver/patches/5/" rel="alternate"/><published>2009-05-04T08:35:21Z</published><updated>2009-05-04T08:35:21Z</updated><author><name>Anthony Giardullo</name><uri>https://sourceforge.net/u/agiardullo/</uri></author><id>https://sourceforge.nete15e74c89c9c174927b0e8659c284b9405119782</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Add a new hashing option for bucket stores that is similar to key_modulo, but uses an additional 'bucket_range' parameter to calculate the modulo.  If you had a 100 buckets and wanted to split them into 2 stores, 'key_modulo' can divide the buckets based on whether they&lt;br /&gt;
're odd or even (ie mod 2).  'key_range' can be used if you want to group the first 50 buckets and the second 50 buckets.&lt;/p&gt;
&lt;p&gt;This is an extra option and shouldn't affect anything else.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>perl scribe namespace</title><link href="https://sourceforge.net/p/scribeserver/patches/4/" rel="alternate"/><published>2009-04-29T05:43:56Z</published><updated>2009-04-29T05:43:56Z</updated><author><name>Jon Schutz</name><uri>https://sourceforge.net/u/jonschutz/</uri></author><id>https://sourceforge.net0f2e4a190c52b9ecfa15eb82304ad1fa4b398c40</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Currently there is no namespace set in if/scribe.thrift for perl, resulting in generation of Constants.pm and Types.pm at root level, leaving them practically uninstallable/unusable due to potential for name conflict with other files.&lt;/p&gt;
&lt;p&gt;The attached patch adds 'namespace perl Scribe.Thrift' to move packages into the Scribe::Thrift namespace.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>store improvements</title><link href="https://sourceforge.net/p/scribeserver/patches/3/" rel="alternate"/><published>2009-02-24T00:36:43Z</published><updated>2009-02-24T00:36:43Z</updated><author><name>Anthony Giardullo</name><uri>https://sourceforge.net/u/agiardullo/</uri></author><id>https://sourceforge.net61b329da553309fd02534efe7c962ccafdcad1af</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Fixed file stores to rotate to a new file after the file size exceeds the max_size(if max_size is set).  Previously, many messages could be written before checking whether the file size has exceeded max_size.&lt;br /&gt;
&lt;/p&gt;
&lt;p&gt;Cleaned up some inconsistencies in re-establishing new connections on failure in conn_pool.cpp/store.cpp when using connection pooling.&lt;/p&gt;
&lt;p&gt;Added the ability to specify bucket definitions explicitly when using a bucketStore:&lt;/p&gt;
&lt;p&gt;#Explicitly&lt;br /&gt;
&amp;lt;store&amp;gt;&lt;br /&gt;
type=bucket&lt;br /&gt;
num_buckets=2&lt;br /&gt;
bucket_type=key_hash&lt;/p&gt;
&lt;p&gt;&amp;lt;bucket0&amp;gt;&lt;br /&gt;
...&lt;br /&gt;
&amp;lt;/bucket0&amp;gt;&lt;/p&gt;
&lt;p&gt;&amp;lt;bucket1&amp;gt;&lt;br /&gt;
...&lt;br /&gt;
&amp;lt;/bucket1&amp;gt;&lt;/p&gt;
&lt;p&gt;&amp;lt;bucket2&amp;gt;&lt;br /&gt;
...&lt;br /&gt;
&amp;lt;/bucket2&amp;gt;&lt;br /&gt;
&amp;lt;/store&amp;gt;&lt;/p&gt;
&lt;p&gt;#Implicitly&lt;br /&gt;
&amp;lt;store&amp;gt;&lt;br /&gt;
type=bucket&lt;br /&gt;
num_buckets=2&lt;br /&gt;
bucket_type=key_hash&lt;/p&gt;
&lt;p&gt;&amp;lt;bucket&amp;gt;&lt;br /&gt;
...&lt;br /&gt;
&amp;lt;/bucket&amp;gt;&lt;br /&gt;
&amp;lt;/store&amp;gt;&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>scribe server improvements</title><link href="https://sourceforge.net/p/scribeserver/patches/2/" rel="alternate"/><published>2009-02-24T00:26:27Z</published><updated>2009-02-24T00:26:27Z</updated><author><name>Anthony Giardullo</name><uri>https://sourceforge.net/u/agiardullo/</uri></author><id>https://sourceforge.netc472e5d6456c62f33a9abae8c106340d1a3e26eb</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Added support to use multiple threads to process incoming Thrift connections.  This is controlled by the configuration option “num_thrift_server_threads”.  The default value is 1, which is the same as the current behavior. &lt;/p&gt;
&lt;p&gt;Added support to be able to reconfigure Scribe with a new config file without stopping Scribe.  This is implemented in scribeHandler::reinitialize().  This function can be called using scribe_ctrl script with the command ‘reload’.&lt;/p&gt;
&lt;p&gt;Added support to configure multiple stores for different categories using a single store definition.  In the config file, you can now specify a whitespace-delimited list of categories using the ‘categories=’ in place of ‘category=’.&lt;/p&gt;&lt;/div&gt;</summary></entry></feed>