==================================================
Changes from 2016-11-16 00:00:00 +0000 to present.
==================================================
------------------------------------------
version 0.010 at 2018-11-16 19:39:31 +0000
------------------------------------------
Change: 9e8c423eb30bf62d10ad4d49abea743b2aefa94e
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-11-16 11:36:54 +0000
Update Perl::Critic policies to pass tests
Change: 2cbed9b55c3d37253edab943ecf329537e2a3e23
Author: Brad Lhotsky <blhotsky@craigslist.org>
Date : 2018-11-16 11:33:33 +0000
Set more reasonable defaults for timeouts, failures, and batch recovery
Change: 2eec820e0d87e4805b93e7a2bdaec39cc73e037e
Author: Brad Lhotsky <blhotsky@craigslist.org>
Date : 2018-11-15 21:52:59 +0000
Fixup how failures are handled
New options:
* MaxFailedRatio: Controls which percentage of bulk requests can fail
each StatsInterval before we back off the cluster * BacklogInterval: How
frequently we attempt to flush
MaxReceoveryBatches from the disk * CleanupInterval: How frequently we
evaluate BatchDiskSpace and discard
batches
Incorporate those mechanism into the process and squash a bug that
prevented BatchDiskSpace from being evaluated properly.
------------------------------------------
version 0.009 at 2018-11-15 05:45:47 +0000
------------------------------------------
Change: f2961717b9be9e4e83e9dfbe68f97f384883f72b
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-11-14 21:45:47 +0000
Fix bug in the cleanup size calculations
I got too clever and tried to do something in one statement that should be
two.
Change: f48972458c836fb4ae10455f9e4ce5351c1a0947
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-11-14 21:45:06 +0000
Stop words added for new options
------------------------------------------
version 0.008 at 2018-11-15 05:38:24 +0000
------------------------------------------
Change: 416da63d80f582a34fd18e72b2bf26b7cd5c4b97
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-11-15 06:38:24 +0000
Version bump to 0.008
Change: 3cbb85abab58e255afdd90384174213b89dbfa7d
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-11-15 06:37:31 +0000
Remove fatal error
I was attempting to add more debugging information, but inadvertently added
a fatal exception.
------------------------------------------
version 0.007 at 2018-11-15 05:20:09 +0000
------------------------------------------
Change: 995f86af18dd66327f2fddd7c45dda79f81990ce
Author: Brad Lhotsky <blhotsky@craigslist.org>
Date : 2018-11-14 21:20:09 +0000
Fix the bug with the diskbased queues
In the previous release, I failed to restore the batch content into the
heap's batch tracking before calling the save event. This led to the data
simpyl being discarded, which is not what was intended.
------------------------------------------
version 0.006 at 2018-11-13 23:14:55 +0000
------------------------------------------
Change: d1532374a1693ce8d0665ab3c08294104d31cab6
Author: Brad Lhotsky <blhotsky@craigslist.org>
Date : 2018-11-13 15:14:55 +0000
Fix typo in the name of the pending_requests_count event
------------------------------------------
version 0.005 at 2018-11-13 04:22:56 +0000
------------------------------------------
Change: 6c742a332a7c58d8357424efa3dbe2f264d77cca
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-11-12 20:22:56 +0000
Track pending_requests to the connection pool
Timeouts are causing unexpected failures, but we might be able to detect
them by detecting high number of pending requests. Default the
MaxPendingRequests to 100 for this first exploration.
------------------------------------------
version 0.004 at 2018-05-03 18:31:11 +0000
------------------------------------------
Change: b8090c274d0dad1f31504f7e147875378d2c2c84
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-05-03 11:31:11 +0000
Documentation updated for v0.004
Change: 83124602857e4d52dde358cbf3c59d8d6d0cbf4c
Author: Brad Lhotsky <blhotsky@craigslist.org>
Date : 2018-05-03 09:39:49 +0000
Return the created POE::Session object
------------------------------------------
version 0.003 at 2018-05-03 04:58:25 +0000
------------------------------------------
Change: 27a4afd11e47ba2bdbdb174179e6af013203dc6e
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-05-02 21:58:25 +0000
Don't track *.yaml files used for testing
Change: deb26ad8c5e9c48a29de7d7b9d5ae8dc9b757539
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-05-02 21:55:44 +0000
Release 0.003 Ready
* Fix race condition involving uninitialized $heap->{queue} * Allow control
of the connection pool via MaxConns{PerServer,Total} * Remove the template
stubs and docs, this is not the place for managing
templates.
Change: dbf5f49034f39d2389421bdc8b503f3397ff6f36
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-04-30 20:12:51 +0000
Documentation updates
Change: 79d5bcb45b5c2dc9752e88059b417a14b5e77e64
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-04-25 14:23:32 +0000
Disable distributing share files
------------------------------------------
version 0.002 at 2018-04-25 21:05:40 +0000
------------------------------------------
Change: 3ba4bfa2a3b7a04aa3c8aa03b7c3ce285a757eff
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-04-25 14:05:40 +0000
Add documentation to the file-to-elasticsearch.pl script
Change: e23668fa086fe3e6e8dfaa4c6804cc0a930e4e48
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-04-25 13:48:55 +0000
Release version 0.002
* Fix StatsHandler * ship file-to-elasticsearch.pl which provides a sample
implementation
Change: 2aa98b914c765cb6c149886e27ab50024c26578b
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-04-25 12:05:24 +0000
Process a shutdown event.
Change: e08cce392ee63cb958ef4cd8d59f2aeb1107a08b
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-03-05 16:32:35 +0000
Enable the split extracter's "from" parameter.
Change: f60872eb6db203e977e5fdcb93d10e8622ee5b09
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-03-05 11:21:58 +0000
Allow _index to contain a strftime compatible string
Change: 39ce4995e73016d9749d88771c2a4a3aeb1701a3
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-03-05 10:30:29 +0000
Make it prettier when there's nothing to report.
Change: cc66b59afdd34fc11d439fccad500190c7547408
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-03-04 12:16:03 +0000
Fixes for the file-to-elasticsearch.pl script
This enables logging, stats collections, records the source of the event in
the _path attribute, and allows the user to specify a different index/type
for each log file.
Change: 0227a1c0a3a8662235113600d8110da032141ec5
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-03-02 23:24:45 +0000
Fixing compile errors on the example implementation
Change: 237c52dabaeba87a9bfcbe551a3090d57226e8b8
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-03-02 23:16:03 +0000
Example implmentation of filebeats/logstash style flat file to
elasticsearch shipper
Change: 4b260fe789b512a31b0c945f297391905c93a243
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-03-02 16:42:06 +0000
Servers Parameter documented
------------------------------------------
version 0.001 at 2018-02-23 05:59:53 +0000
------------------------------------------
Change: 5b1928e1e8ce911c8bc6bf14e4635da8115b50ee
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-22 21:59:53 +0000
Simpler queue/flush logic
Most of the time, simpler is better. Stop trying to do weird stuff and just
do the simplest possible thing to reschedule the queue flushes.
Change: c5902dfe4fa393504311dd711fe9d895d6a549f7
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-22 21:48:14 +0000
Dispatch the correct HTTP response event.
Change: d1c89dd751f4bf911f5067afd88da6f0a6b6dee5
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-22 21:43:16 +0000
Set the alias correctly.
Change: ba11985f66ac2d44b282a404ed2cf0b76b25ed23
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-22 21:36:27 +0000
Ensure the queue method matches the documentation.
Change: 7c4b03165823fbbae659a3f90a7a8c1be5ac2275
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-22 21:31:52 +0000
Comment out unimplemented features.
Change: 5634cba4190c3b8d8bec6fe19cb330bbdc6e5b58
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-22 21:16:00 +0000
Fix accidental deletion of a character.
Change: 9a1eede4fe2bc8ef4dd838a8b5c619c3c60bdfb8
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-22 21:14:01 +0000
Fix the Servers parameter.
While converting this code, I mistakenly overlooked the connection
settings.
Change: 2394ba7a48e3b27fc4b01dfd2484898c73a596cf
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-22 20:44:27 +0000
Allow sending the entire bulk operation as a scalar
The queue mechanism now takes the following:
1. An array reference, in which all elements must be one of: 2. A hash
reference, which is converted into the newline delimited JSON
expected by the Bulk API. 3. An object reference, which must implement an
as_bulk() method which
returns the newline delimited JSON expected by the Bulk API 4. A string of
newline delimited JSON which is passed unchanged to the
Bulk API.
Change: 0308c7f788e07c5e696ea54cbd5e7c4e4d6556c8
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-22 19:24:33 +0000
Documentation updates
Change: 7ddb91c2fbaacafb7ae98a4afb98dce1c57ea72f
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-22 19:11:02 +0000
Document the BatchDiskSpace argument
Change: 78b001c5f55a92e89614553059c5081b7d9ea08a
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-16 23:18:43 +0000
Throttle disk space and allow a shutdown event to be passed along to
prevent recurring tasks from being rescheduled.
Change: 8165e90c16582e936d1ef78c87cb824d8f553890
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-16 21:08:30 +0000
Adjust log levels and how flushes are scheduled
Log4perl's log levels include "TRACE" below "DEBUG" and no "NOTICE" so
adjust log levels to compensate.
Only reschedule the flush routine if there was data. Have the queue event
schedule the flush when it happens.
Change: bdcd29a6b27e01ce7166e403437b7fb43617f8c8
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-16 20:22:26 +0000
Start of the Template Saving
Also, splitting out the batch save job to make it configurable.
Change: 5bbaaa3be95c891744a1d620ba0e7044a20e4f98
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-07 13:08:40 +0000
Allow blessed objects to be sent to the bulk queue as long as they expose a
as_bulk() method
Change: ab526468d4a233dce8f57dbbe01eefe18c21aa3e
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-03 22:32:18 +0000
Run ExtraTests always and load the keepalive PoCo
Change: 11a89c20537f11bc8788246b9640b20a2c810bf2
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-03 22:26:27 +0000
Fix tabs that somehow made it into the documentation
Change: 3251f56c2450e2c0fb37cfde7e541fbba3391acb
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-03 22:22:42 +0000
Preparing for release
* Pod::Coverage tests sated * Example use case provided
Change: b8fe6bfbac39afd771b7c33fa4480a8629ec1582
Author: Brad Lhotsky <brad@divisionbyzero.net>
Date : 2018-02-02 23:15:59 +0000
Initial checkin of the PoCo::ES::Indexer
Change: cffce9ddea13d57df5248fe5afc691050feb0e97
Author: Brad Lhotsky <brad.lhotsky@gmail.com>
Date : 2012-12-22 11:33:13 +0000
Distzilla setup of the module with some basic information.
================
End of releases.
================