[
https://issues.apache.org/jira/browse/COUCHDB-248?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12772829#action_12772829
]
Adam Kocoloski commented on COUCHDB-248:
----------------------------------------
Hi Paul, I don't know what's "normal". As Dip pointed out, SASL ships with a
few event handlers. We use the default one (sasl_report_tty_h) and set
-sasl errlog_type error
in the startup script so that it only logs error-level messages. I think that
logger only handles crash and supervisor reports, becuase we define ?LOG_DEBUG
and ?LOG_INFO to send info_reports and ?LOG_ERROR to send error reports of
specific types (couch_debug/info/error), and those types are only handled by
custom event handlers.
I don't think there's anything too weird about writing our own event handlers
for these messages. The event handlers that ship with SASL are probably much
too verbose for an access log anyway.
One area where we do need to improve is in trimming down those crash reports.
There's a lot of duplicate info in there, in part because our handler and
sasl_report_tty_h sometimes end up effectively logging two copies of the same
crash.
> The log process should use SASL for logging
> -------------------------------------------
>
> Key: COUCHDB-248
> URL: https://issues.apache.org/jira/browse/COUCHDB-248
> Project: CouchDB
> Issue Type: Improvement
> Components: Database Core
> Affects Versions: 0.9
> Environment: any
> Reporter: Ulises Cervino Beresi
> Priority: Minor
>
> As it stands the log process is ad-hoc since it was written from scratch.
> Making it a wrapper around SASL (http://erlang.org/doc/apps/sasl/index.html)
> would be a good improvement.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
