cloudstack-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF GitHub Bot (JIRA)" <>
Subject [jira] [Commented] (CLOUDSTACK-8645) Make better use of different loglevels like INFO, WARN and ERROR instead of DEBUG
Date Mon, 14 Sep 2015 12:46:46 GMT


ASF GitHub Bot commented on CLOUDSTACK-8645:

Github user DaanHoogland commented on a diff in the pull request:
    --- Diff: plugins/hypervisors/kvm/src/com/cloud/hypervisor/kvm/storage/
    @@ -877,8 +877,8 @@ public boolean deletePhysicalDisk(String uuid, KVMStoragePool pool,
    -                s_logger.debug("Succesfully unprotected and removed any snapshots of
" + pool.getSourceDir() + "/" + uuid +
    -                        " Continuing to remove the RBD image");
    +      "Succesfully unprotected and removed all snapshots " +
snaps.size() + " of " + pool.getSourceDir() + "/" + uuid +
    --- End diff --
    This is going to read something like "Succesfully unprotected and removed all snapshots
5 of /<somepath>/<some uuid> Continuing to remove the RBD image", in the logs.
the part "all snapshots # of ..." seems not what you'd want to read. How about "Succesfully
unprotected and removed any remaining snapshots (counting " + snaps.size() + ") of " + ...?

> Make better use of different loglevels like INFO, WARN and ERROR instead of DEBUG
> ---------------------------------------------------------------------------------
>                 Key: CLOUDSTACK-8645
>                 URL:
>             Project: CloudStack
>          Issue Type: Improvement
>      Security Level: Public(Anyone can view this level - this is the default.) 
>    Affects Versions: Future
>         Environment: All CloudStack components
>            Reporter: Wido den Hollander
>              Labels: debug, log4j, logging, loglevel
>             Fix For: Future
> Currently most of the code uses s_logger.debug() for logging events.
> Running on logging level INFO or WARN makes CloudStack logs almost unusable since almost
no interesting information comes by.
> We should improve the logging on various points by replacing DEBUG with INFO, WARN or
ERROR to what seems appropriate.
> This should be done continuously, this isn't something which can be fixed with one big
patch, commit or PR.
> While working on the code you'll spot these issues and can fix them when needed.
> On big environments for example the management server log in level DEBUG becomes almost
unusable due to the high volume of messages coming by.
> There is certainly more to improve on logging, but this is a start.

This message was sent by Atlassian JIRA

View raw message