Automatic merge from master to 5.x branch #3210
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
* remove init collecting of processes * add changelog entry
I reordered the options based on importance (I put the optional config setting at the end). And I changed the wording to further clarify that the `json.message_key` setting is optional. Fixes elastic#2864
* Update docker files to the last major with the most recent minor and bugfix version * Renamed files to Dockerfile-2x to not have to be renamed every time a new bugfix is released * Remove scripts and config files which are not needed anymore To run testsuite for 2x releases, run: `TESTING_ENVIRONMENT=2x make testsuite`
- Update go-ucfg - add support for parsing lists/dictionaries from environment variables and via `-E` flag
…c#3086) Part of elastic#2629. The name of the field was changed, but not in the dashboard. (cherry picked from commit e271d9f)
Update kafka broker query - Switch paritition metricset from client to broker - on connect try to find the broker id (address must match advertised host). - check broker is leader before querying offsets - query offsets for all replicas - remove 'isr' from event, and replace with boolean flag `insync_replica` - replace `replicas` from event with per event `replica`-id - update sarama to get offset per replica id
* Update data.json
…lastic#3096) Remove example for publish_async from the docs
There was a test that was loading a mock template, and this template was assuming 5.x.
virtualenv creates symlinks so `make setup` fails when ran on a network mounted fs. `--always-copy` copies files to the destination dir rather than symlinking.
This prefix is need to run tests with different environments in parallel so one does not affect the other. Like this 2x and snapshot builds should be able to coexist
When decoding a UTF16 string contained in a buffer larger than just the string, more space was allocated than required.
```
BenchmarkUTF16BytesToString/simple_string-4 2000000 846 ns/op 384 B/op 3 allocs/op
BenchmarkUTF16BytesToString/larger_buffer-4 2000000 874 ns/op 384 B/op 3 allocs/op
BenchmarkUTF16BytesToString_Original/simple_string-4 2000000 840 ns/op 384 B/op 3 allocs/op
BenchmarkUTF16BytesToString_Original/larger_buffer-4 1000000 3055 ns/op 8720 B/op 3 allocs/op
```
```
PS C:\Gopath\src\github.com\elastic\beats\winlogbeat> go test -v github.com/elastic/beats/winlogbeat/eventlog -run ^TestBenchmarkBatchReadSize$ -benchmem -benchtime 10s -benchtest
=== RUN TestBenchmarkBatchReadSize
--- PASS: TestBenchmarkBatchReadSize (68.04s)
bench_test.go:100: batch_size=10, total_events=20000, batch_time=5.682627ms, events_per_sec=1759.7494961397256, bytes_alloced_per_event=44 kB, total_allocs=4923840
bench_test.go:100: batch_size=100, total_events=30000, batch_time=53.850879ms, events_per_sec=1856.9799018508127, bytes_alloced_per_event=44 kB, total_allocs=7354285
bench_test.go:100: batch_size=500, total_events=25000, batch_time=271.118774ms, events_per_sec=1844.2101689350366, bytes_alloced_per_event=43 kB, total_allocs=6125665
bench_test.go:100: batch_size=1000, total_events=30000, batch_time=558.03918ms, events_per_sec=1791.9888707455987, bytes_alloced_per_event=43 kB, total_allocs=7350324
PASS
ok github.com/elastic/beats/winlogbeat/eventlog 68.095s
PS C:\Gopath\src\github.com\elastic\beats\winlogbeat> go test -v github.com/elastic/beats/winlogbeat/eventlog -run ^TestBenchmarkBatchReadSize$ -benchmem -benchtime 10s -benchtest
=== RUN TestBenchmarkBatchReadSize
--- PASS: TestBenchmarkBatchReadSize (71.85s)
bench_test.go:100: batch_size=10, total_events=30000, batch_time=5.713873ms, events_per_sec=1750.1264028794478, bytes_alloced_per_event=25 kB, total_allocs=7385820
bench_test.go:100: batch_size=100, total_events=30000, batch_time=52.454484ms, events_per_sec=1906.4147118480853, bytes_alloced_per_event=24 kB, total_allocs=7354318
bench_test.go:100: batch_size=500, total_events=25000, batch_time=260.56659ms, events_per_sec=1918.8952812407758, bytes_alloced_per_event=24 kB, total_allocs=6125688
bench_test.go:100: batch_size=1000, total_events=30000, batch_time=530.468816ms, events_per_sec=1885.124949550286, bytes_alloced_per_event=24 kB, total_allocs=7350360
PASS
ok github.com/elastic/beats/winlogbeat/eventlog 71.908s
```
…tal-batch-read-size Document batch_read_size is experimental in Winlogbeat
When reading a batch of large event log records the Windows function
EvtNext returns errno 1734 (0x6C6) which is RPC_S_INVALID_BOUND ("The
array bounds are invalid."). This seems to be a bug in Windows because
there is no documentation about this behavior.
This fix handles the error by resetting the event log subscription
handle (so events are not lost) and then retries the EvtNext call
with maxHandles/2.
Fixes elastic#3076
Currently fetching container stats is very slow as each request takes up to 2 seconds. To improve the fetching time if lots of containers are around, this creates the rrequests in parallel. The main downside is that this opens lots of connections. This fix should only temporary until the bulk api is available: moby/moby#25361
Add beat name to project namespace * This makes sure different beats environment do not affect each other for example when Kafka is used * It also allows to run the testsuites of all the beats in parallel Introduce `stop-environment` command to stop all containers
* Add doc for decode_json_fields processor * Use changed param names * Add example of decode_json_fields processor * Fix intro language about processors
I created a basic version of amazonbeat, which reads data from an amazon product periodically. This beat does not yet publish to elasticsearch.
Previously the data was read into a []byte encoded as UTF16. Then that data was converted to []uint16 so that we can use utf16.Decode(). Then the []rune slice was converted to a string which did another data copy. The XML was unmarshalled from the string. This PR changes the code to convert the UTF16 []byte directly to UTF8 and puts the result into a reusable bytes.Buffer. The XML is then unmarshalled directly from the data in buffer. ``` BenchmarkUTF16ToUTF8-4 2000000 1044 ns/op 4 B/op 1 allocs/op ``` ``` git checkout 6ba7700 PS > go test github.com/elastic/beats/winlogbeat/eventlog -run TestBenc -benchtest -benchtime 10s -v === RUN TestBenchmarkBatchReadSize --- PASS: TestBenchmarkBatchReadSize (67.89s) bench_test.go:100: batch_size=10, total_events=30000, batch_time=5.119626ms, events_per_sec=1953.2676801000696, bytes_alloced_per_event=44 kB, total_allocs=7385952 bench_test.go:100: batch_size=100, total_events=30000, batch_time=51.366271ms, events_per_sec=1946.802795943665, bytes_alloced_per_event=44 kB, total_allocs=7354448 bench_test.go:100: batch_size=500, total_events=25000, batch_time=250.974356ms, events_per_sec=1992.2354138842775, bytes_alloced_per_event=43 kB, total_allocs=6125812 bench_test.go:100: batch_size=1000, total_events=30000, batch_time=514.796113ms, events_per_sec=1942.5166094834128, bytes_alloced_per_event=43 kB, total_allocs=7350550 PASS ok github.com/elastic/beats/winlogbeat/eventlog 67.950s git checkout 833a806 (elastic#3113) PS > go test github.com/elastic/beats/winlogbeat/eventlog -run TestBenc -benchtest -benchtime 10s -v === RUN TestBenchmarkBatchReadSize --- PASS: TestBenchmarkBatchReadSize (65.69s) bench_test.go:100: batch_size=10, total_events=30000, batch_time=4.858277ms, events_per_sec=2058.3429063431336, bytes_alloced_per_event=25 kB, total_allocs=7385847 bench_test.go:100: batch_size=100, total_events=30000, batch_time=51.612952ms, events_per_sec=1937.49816906423, bytes_alloced_per_event=24 kB, total_allocs=7354362 bench_test.go:100: batch_size=500, total_events=25000, batch_time=241.713826ms, events_per_sec=2068.561853801445, bytes_alloced_per_event=24 kB, total_allocs=6125757 bench_test.go:100: batch_size=1000, total_events=30000, batch_time=494.961643ms, events_per_sec=2020.3585755431961, bytes_alloced_per_event=24 kB, total_allocs=7350474 PASS ok github.com/elastic/beats/winlogbeat/eventlog 65.747s This PR (elastic#3118) PS > go test github.com/elastic/beats/winlogbeat/eventlog -run TestBenc -benchtest -benchtime 10s -v === RUN TestBenchmarkBatchReadSize --- PASS: TestBenchmarkBatchReadSize (65.80s) bench_test.go:100: batch_size=10, total_events=30000, batch_time=4.925281ms, events_per_sec=2030.341009985014, bytes_alloced_per_event=14 kB, total_allocs=7295817 bench_test.go:100: batch_size=100, total_events=30000, batch_time=48.976134ms, events_per_sec=2041.8108134055658, bytes_alloced_per_event=14 kB, total_allocs=7264329 bench_test.go:100: batch_size=500, total_events=25000, batch_time=250.314316ms, events_per_sec=1997.4886294557757, bytes_alloced_per_event=14 kB, total_allocs=6050719 bench_test.go:100: batch_size=1000, total_events=30000, batch_time=499.861923ms, events_per_sec=2000.5524605641945, bytes_alloced_per_event=14 kB, total_allocs=7260400 PASS ok github.com/elastic/beats/winlogbeat/eventlog 65.856s ```
Mongobeat discovers instances in a mongo cluster and can be configured to ship multiple document types - from the commands db.stats() and db.serverStatus()
* Merge snapshot and latest build for Logstash into 1 docker file
* Pass certificate options to import dashboards script -cert for client certificate -key for client certificate key -cacert for certificate authority
Loading and creating docker images takes quite a bit of time on the travis builds. Especially calls like apt-get update and install take lots of time and bandwidth and fail from time to time, as a host is not available. Following actions were taken: * Fake Kibana container is now based on alpine * Redis stunnel container was also switched to alpine
The enabled config allows easily to enable and disable a specific prospector. This is consistent with metricbeat where each modules has an enabled config. By default enabled is set to true.
Contains the Nginx module, including the fields.yml and several pipelines.
Added the 5.1.0 and 5.1.1 sections, removed duplicates.
- compare broker names to hostname - try to lookup metricbeat host machine fqdn and compare to broker name - compare all ips of local machine with resolved broker name ips
* Contains slowlog and errors filesets * Test files for two mysql versions (5.5 and 5.7) * Add support for built-in variables (e.g. `builtin.hostname`) * Contains a sample Kibana dashboard Part of elastic#3159.
Add a new Makefile rule: fix-permissions fix-permissions runs a docker container that changes the ownership of all files from root to the user that runs the Makefile
* packer: Enable overriding of vendor and license * packer: customize URL of documentation link * packer: location of readme.md.j2 folder can be specified with PACKER_TEMPLATES_DIR
* Basic parsing of syslog fields * Supports multiline messages if the lines after the first one start with a space. * Contains a simple Kibana dashboard
We generally need more than one path per OS, because the logs location is not always the same. For example, depending on the linux distribution and how you installed it, MySQL can have it's error logs in a number of default "paths". The solution is to configure them all, which means that Filebeat might try to access unexisting folders. This also improves the python prototype to accept multiple modules and to accept namespaced parameters. E.g.: ./filebeat.py --modules=nginx,syslog -M nginx.access.paths=...
…c#3193) - re-use common.LocalIPAddrs in partition module for resolving IPs - add missing net.IPAddr type switch to common.LocalIPAddrs - update matching to extract addresses early on using strings.ToLower => ensure case insensitive matching by lowercasing
(cherry picked from commit 0f4103f)
|
jenkins, test it |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Doing a merge now because I'd like the 5.x branch to be in a known state, so we can configure the snapshots for Heartbeat.
There were no conflicts.