You are viewing a plain text version of this content. The canonical link for it is here.
Posted to notifications@skywalking.apache.org by GitBox <gi...@apache.org> on 2018/05/28 00:56:52 UTC
[GitHub] zpf1989 opened a new issue #1271: can sw collector start when
elasticsearch is not green?
zpf1989 opened a new issue #1271: can sw collector start when elasticsearch is not green?
URL: https://github.com/apache/incubator-skywalking/issues/1271
Hi,firstly thank you for your great job!!!
Here is my case:
I'm using elasticsearch as storage.
I start es firstly.
When it prints "started",I start sw collector then,but its console close instantly,and below is the log:
_org.apache.skywalking.apm.collector.core.module.ModuleStartException: **Field named mi in application_alarm_list_month index not found**.
at org.apache.skywalking.apm.collector.storage.es.StorageModuleEsProvider.start(StorageModuleEsProvider.java:300) ~[collector-storage-es-provider-5.0.0-beta.jar:5.0.0-beta]
at org.apache.skywalking.apm.collector.core.module.BootstrapFlow.start(BootstrapFlow.java:61) ~[apm-collector-core-5.0.0-beta.jar:5.0.0-beta]
at org.apache.skywalking.apm.collector.core.module.ModuleManager.init(ModuleManager.java:68) ~[apm-collector-core-5.0.0-beta.jar:5.0.0-beta]
at org.apache.skywalking.apm.collector.boot.CollectorBootStartUp.main(CollectorBootStartUp.java:45) [apm-collector-boot-5.0.0-beta.jar:5.0.0-beta]
Caused by: org.apache.skywalking.apm.collector.storage.StorageInstallException: Field named mi in application_alarm_list_month index not found.
at org.apache.skywalking.apm.collector.storage.es.base.define.ElasticSearchStorageInstaller.columnCheck(ElasticSearchStorageInstaller.java:72) ~[collector-storage-es-provider-5.0.0-beta.jar:5.0.0-beta]
at org.apache.skywalking.apm.collector.storage.StorageInstaller.install(StorageInstaller.java:60) ~[collector-storage-define-5.0.0-beta.jar:5.0.0-beta]
at org.apache.skywalking.apm.collector.storage.es.StorageModuleEsProvider.start(StorageModuleEsProvider.java:298) ~[collector-storage-es-provider-5.0.0-beta.jar:5.0.0-beta]
... 3 more_
——————————————————————————————————————————————
I think it may be caused by the state of es.
So I restart es again ,and wait (**about 1 minutes**).
After it prints " Cluster health status changed from [RED] to [GREEN]",like below:
I start sw collector,and it's OK.
I don't know why es's state from yellow to green costs more than 1 minutes....
**But It seems that sw collector can't start when es in required state.**
And I think it's not friendly for auto deploy,especially in container,because it's hard to determine when the state of es is OK for sw collector.
——————————————————————————————————————————————
**So,could you guys make some retry policies when sw collector start?
Or if it is possible to start sw collector without es's state in green?**
——————————————————————————————————————————————
Below is my system:
skywalking-beta for windows;
elasticsearch 5.6.6
skywalking-aspnetcore 0.2.0
Thank you,master!!!
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
With regards,
Apache Git Services