Commit Graph

122 Commits

Author SHA1 Message Date
吴晟 Wu Sheng aaa403ec00
Set telemetry module and APIs (#2123)
* Finish the APIs of telemetry module.

* Add prometheus license.

* Active default telemetry provider.
2019-01-07 09:48:53 +08:00
langyan1022 6a4e7ffc6c fix missing ojdbc component id in component-libraries.yml (#2122)
* add ojdbc component id
2019-01-04 20:56:38 +08:00
吴晟 Wu Sheng ec391becad
Provide two start up modes and documents. (#2116) 2019-01-03 15:19:45 +08:00
吴晟 Wu Sheng 95214aecf3
update version for next iteration. (#2084) 2018-12-23 09:14:14 +08:00
吴晟 Wu Sheng c592aea8db Refactor register API for making integration easier. (#2075)
* Refactor register API for making integration easier.

* Fix wrong refactor.
2018-12-21 22:32:06 +08:00
吴晟 Wu Sheng c7884694da
Move node type from network address to service level[Need storage module update] (#2051)
* Change srcLayer to nodeType. Add nodeType in service inventory entity.

* NodeType value refactor.

* Format codes.

* Fix compile issue.

* Fix bugs, and make query on service inventory instead of network inventory
2018-12-17 09:05:16 -08:00
彭勇升 pengys dfe2b08599 [OAP Server][Cluster Plugin] Provide consul cluster plugin (#2036)
* Provide consul cluster plugin.

* Change log and application config file.

* Add consul's config item into the docker's application.yml.

* Fixed bug.

* Add consul description into  readme.

* Add okhttp component license.
2018-12-12 08:14:41 -08:00
彭勇升 pengys 7532583d90 Endpoint name need fuzzy query. (#2026) 2018-12-09 13:01:58 +08:00
吴晟 Wu Sheng 40eb651238
OAL tool module established (#2015)
* Move oal tool back into main repo, and fix compile issue.

* Fix two bugs.

* Remove all generated codes.

* Make all oal generated codes out of source control.

* Make mvn compile and package work.

* Fix test cases.

* Make compile right and provide new documents.

* Remove an unnecessary tokens file.
2018-12-07 09:31:16 +08:00
彭勇升 pengys 27c475f74c 1. Make storage and cluster using same name space. (#2014)
2. Zookeeper implementation of the cluster module using name space to support using same zookeeper cluster by different OAP clusters.
2018-12-06 21:03:00 +08:00
吴晟 Wu Sheng f2b158862c
[OAL refactor step 1]Refactor dispatcher manager (#2006)
* Remove the hard codes about DispatcherManager.

* Add comments.
2018-12-06 14:40:29 +08:00
吴晟 Wu Sheng 5f5be77601
Fix the alarm being started too many times. (#2003)
* Fix the alarm being started too many times.

* Remove unnecessary interface.

* Fix wrong status changed
2018-12-05 20:56:08 +08:00
吴晟 Wu Sheng 44a85a4f5a Support service instance metric. (#1988) 2018-12-01 22:57:28 +08:00
吴晟 Wu Sheng 2d8f390351 Make endpoint and instance alarm active (#1987)
* Support alarm in service instance and endpoint.

* Fix alarm default settings and document.

* Fix wrong default setting

* Make service instance and endpoint scopes available in alarm core.

* Fix alarm bug.
2018-12-01 14:18:31 +08:00
彭勇升 pengys 76d2fa34e0 Parse segment binary with version 1 prototype by default. (#1980) 2018-11-30 13:15:04 +08:00
吴晟 Wu Sheng 215eab7428
Mysql style storage implementation (#1964)
Also fixed several bugs in this merge.
2018-11-28 22:59:02 +08:00
彭勇升 pengys fd47fcb830 Remote clients selector error: / by zero (#1949)
* Fixed the bug of hash selector error cause of remote clients not build.

#1948

* Fixed a style check error.
2018-11-22 22:28:36 +08:00
彭勇升 pengys 7c8a683c2f
gRPC client usage improve (#1946)
* Close the clients which are unreachable.
Remote client manager test case and comments.

* Test the GRPCRemote client.

* 1. Catch the throwable for onComplete method cause of it will throw exception when connection lost.
2. Check the gRPC channel state, send message when state is ready, wait 5 second when state is not ready.

Notice: gRPC channel getState with true parameter will trigger reconnect operation.

* gRPC client will reconnect to the server when network recorvered.

* Recovery application.yml

* Recovery proto module commit id.

* no message

* Fixed compile error.
2018-11-22 20:50:27 +08:00
吴晟 Wu Sheng c6ada8c98f
Support new v2 protocol and make concept consistently (#1937)
* Update new protocol

* Support new protocol at agent side.

* Fix test case.

* Make backend supports new trace protocol

* Fix tests.

* Make all new services available in gRPC.

* Fix bugs.

* Finish new document for v2 trace protocol.

* Fix wrong version requirement of zk

* Fix trace query bug.

* Find a client side dictionary bug.

* Fix register bugs.

* Fix break doc links.
2018-11-21 22:33:04 +08:00
彭勇升 pengys afd6164335 OAP server testing module. (#1931)
* OAP server testing module.

#1930

* #Missing delete code.
2018-11-18 20:22:56 +08:00
吴晟 Wu Sheng 6dd98b37bf Extend StorageEntity annotation to support source scope metadata (#1904)
* Add source attr into StorageEntity, and remove the this attr from InventoryType.

* Make RegisterLockInstaller works based on InventoryProcess registered info.
2018-11-12 14:19:16 +08:00
吴晟 Wu Sheng 83e0e5758f Make compare right. (#1899) 2018-11-09 22:23:50 +08:00
ScienJus 926911700e Fix logs with missing variables (#1884) 2018-11-06 18:32:02 +08:00
吴晟 Wu Sheng 72a39e1f28
Sync 6.0.0-alpha release and version up to beta. (#1883)
* Lock oracle and resin plugins version.

* Move resin and oracle plugin to openskywalking

* Restore submodule active.

* Add another submodule active command.

* Remove the wrong test cases in release and fix some doc issues.

* [maven-release-plugin] prepare release v6.0.0-alpha

* [maven-release-plugin] prepare for next development iteration

* Fix some doc
2018-11-06 15:45:15 +08:00
吴晟 Wu Sheng 4399c66278
Make metric value to long to avoid too big integer issue. (#1876)
* Make metric value to long to avoid too big integer issue.

* Update protocol.

* Add a new filter to service number.
2018-11-03 23:25:28 +08:00
彭勇升 pengys 884dd8c59e Add detect point to be a part of unique key of the endpoint id. (#1874) 2018-11-03 00:55:00 +08:00
彭勇升 pengys 6b1014c336 Fixed the collector OOM bug. (#1862)
* Fixed the bug of remote client not blocked when not received on complete message, it will carry the out of a memory exception.

* Sleep 10ms, not to sleep max 10ms.

* No more than 10 stream observers are allowed at the same time to send remote message. Otherwise block the remote queue.

* no message
2018-11-02 10:33:24 +08:00
吴晟 Wu Sheng 07e799e422 Fix bugs (#1854)
* Avoid service inventory object concurrency situation.

* Fix concurrency for minute persistence and others

* Fix document todo.

* Update license. Fix package with dirty agent plugin jar files.

* Fixed not combine error in persistence register worker.
2018-10-31 23:11:52 +08:00
吴晟 Wu Sheng dbf0444ec1
Fix NPE in protobuf and illegal query parameter. (#1847) 2018-10-30 12:17:17 +08:00
吴晟 Wu Sheng b6241b2ad2
Optimize collector settings and indicator (#1841)
* Support maxConcurrentCallsPerConnection and maxMessageSize in grpcServer

* Make percent percent is 1/10000

* Fix test cases.

* Make service instance indicator dispatch based on server side only.

* Set init data right in AlarmMeta.

* Make endpoint query based on server side only.
2018-10-29 20:45:44 +08:00
彭勇升 pengys 948ef6d1c4 Fixed the bugs of the OAP server running in cluster mode. (#1844)
* 1. Fixed the bug of the cluster node sync failure cause by curator instance cache is null.
2. Fixed the bug of the collection in remote data builder is not initialized.

* 1. Re-generate indicator.

* Fixed the bug of the collection in remote data builder is not initialized.

* Fixed the bug of the remote client created but not  initialize.
2018-10-29 17:31:30 +08:00
吴晟 Wu Sheng 974192ecd8 Keep service instance register based on UUID, but change name to a meaning string. (#1836) 2018-10-29 09:28:33 +08:00
吴晟 Wu Sheng 8cd327d8bb
Revert #1790 reset feature and provide default alarm settings. (#1833)
* Revert reset feature.

* Set default dist alarm settings and make dist package setting files right.

* Support empty list.
2018-10-28 22:28:21 +08:00
吴晟 Wu Sheng a9489b7067
H2 storage (#1712)
* H2 storage code skeleton.

* Add batch dao for H2.

* Fix wrong refactor.

* Add codes of other dao. No metric dao, wait for @peng-yongsheng new PR.

* Change license of distribution

* Add topology query in H2.

* Add more daos.

* Add some daos.

* Commit all H2 daos. Wait @peng-yongsheng to submit new IAggregationQueryDAO interface.

* Fix some CIs

* Fix wrong refactor, which causes h2 plugin failure.

* Fix all codes of H2 storage. :) and CI should pass. No test.

* Fix CI and typo

* Fix connection closed.

* Fix register bug

* no message

* Fixed service inventory register bug.

* Fix a compile issue.

* Sync change

* Reset es as default storage.

* Fix rat

* Fix H2 compile issue.

* Fix bugs for H2 storage.

* Fix bug in mesh.

* 1. Fix Thermodynamic query bug
2. Fix service inventory query bug
3. Fix missing traceid in trace query bug

* Fix query bug.

* Submit reverse caused by prev PR.

* Fix duplicate spans.

* Use H2 as default storage implementor.

* Fix bug in topology query.

* Fix CI

* Fix rat

* Fix two query in H2 storage.

* Fix H2 query

* Make release running in H2 mode and develop mode running is ES mode in default
2018-10-28 10:31:58 +08:00
刘新元 Liu XinYuan b11d479337 Add service and instance reset function (#1790)
* Manually trigger agent registration and write registration status to file

* Delete unused variables

* Code optimization, encapsulating some methods

* add clear DataCarrier function

* revert SkyWalkingAgent

* Optimize code and logic, including updating applicatin_id in unRegisterOperationNames

* Server end metadata loss notification

* Server end metadata loss notification

* fix file stream close and other problem

* revert to old

* fix some bug

Signed-off-by: Liu-XinYuan <879928098@qq.com>

* fix some bugs in agent reset

* add close inputStream and fileChannel and fix some bug

* fixed some bugs
2018-10-27 19:58:15 +08:00
彭勇升 pengys 9b3192fbac Set span layer into network inventory for distinguish what the network address is. MQ? DB? ... (#1824) 2018-10-26 22:57:05 +08:00
吴晟 Wu Sheng d9556c38fc
Make Mesh and Istio receivers ready (#1821)
* Make receiver more effective

* Add Istio test case and all source dispatch in

* Refactor mock data

* FIx rat.

* Support call component.

* Fix ThermodynamicIndicator bug.

* Fix test cases.

* Fix missing calculate in db merging.

* Add codes for debug.

* 1. Fixed elasticsearch bulk process not fresh bug. (#1819)

2. Fixed the bug of source register no queue but wait the end for batch tag.

* Remove debug log, and restore TTL timer for real scenarios.
2018-10-26 15:17:33 +08:00
彭勇升 pengys 24e958fec0 1. Fixed the bug of inventory equals method. (#1814)
2. Fixed the bug of agent mock register order.
3. Fixed the bug of all heatmap.
2018-10-25 21:23:20 +08:00
彭勇升 pengys 6c92fc5484 1. Delete service component indicator. (#1809)
2. Make service relation and endpoint relation to record component id.
3. Delete service mapping indicator.
4. Make service inventory to record mapping service id.
5. Refactor global topology query and service topology query.
2018-10-25 11:19:31 +08:00
彭勇升 pengys 7bb577a4e9 Provide auto delete history data service. (#1797) 2018-10-22 20:20:29 +08:00
彭勇升 pengys 4063397988 Trigger service heartbeat when received service instance heartbeat. (#1785)
Fixed some meta query bugs.
2018-10-17 22:44:21 +08:00
彭勇升 pengys 4e99e896c2
Set instance attributes into return value. (#1783)
* Set instance attributes into return value.

* Delete unused code.
2018-10-17 18:40:10 +08:00
吴晟 Wu Sheng cb0b817ffb
Fix indicator merge failue. (#1782) 2018-10-17 16:42:03 +08:00
彭勇升 pengys a12e8badb8 Provide new getEndpointInfo service in query protocol. (#1780) 2018-10-17 12:19:33 +08:00
彭勇升 pengys b367c36db9
Fixed the endpoint topology bug. (#1778)
* Delete the client side endpoint relation indicator.
Fixed the endpoint topology bug.

* Fixed package failure bug.
2018-10-17 11:28:16 +08:00
彭勇升 pengys f78abfb28c Fixed the bug of service topology and global topology. (#1775) 2018-10-16 18:40:42 +08:00
彭勇升 pengys b426390b79
Support metrics downsampling. (#1765)
* Support metrics downsampling.

* Fixed a check style error.
2018-10-15 15:27:38 +08:00
彭勇升 pengys a81b9920cd Set the function attribute for indicator and use AVG function to be the default value. (#1757) 2018-10-12 12:29:22 +08:00
彭勇升 pengys 966d5f54f1 Fixed the bugs, #1750, #1749, #1748, #1746, #1745 (#1755) 2018-10-11 19:37:57 +08:00
彭勇升 pengys 7c192b19d6
Alarm query and alarm persistence test. (#1753)
* Alarm query and alarm persistence test.

* Fixed the package error.
2018-10-11 15:55:04 +08:00