You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@ignite.apache.org by "Nikolay Tikhonov (JIRA)" <ji...@apache.org> on 2015/02/19 10:49:11 UTC

[jira] [Updated] (IGNITE-306) Collect N (e.g. 1000) last errors for grid node

     [ https://issues.apache.org/jira/browse/IGNITE-306?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Nikolay Tikhonov updated IGNITE-306:
------------------------------------
    Attachment: ignite-306.patch

> Collect N (e.g. 1000) last errors for grid node
> -----------------------------------------------
>
>                 Key: IGNITE-306
>                 URL: https://issues.apache.org/jira/browse/IGNITE-306
>             Project: Ignite
>          Issue Type: Bug
>          Components: general
>            Reporter: Nikolay Tikhonov
>            Assignee: Nikolay Tikhonov
>             Fix For: sprint-2
>
>         Attachments: ignite-306.patch
>
>
> Many errors in disco or communication are not output to logs (other than debug), e.g. in case when connection has been reestablished. However, this may affect performance badly. We need to collect last errors for components including:
> time
> thread ID and name
> Message
> Stacktrace
> So in case we have strange performance drops we can check last errors.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)