[
https://issues.apache.org/jira/browse/SOLR-11934?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16348842#comment-16348842
]
Erick Erickson commented on SOLR-11934:
---------------------------------------
bq: if someone tells me when it went wrong, with a default Solr install I have
something to dig into to find out what went wrong
Yeah, that's the tension, isn't it? When it does hit the fan you want
_everything_, and it's awkward to say "Sorry, the info we need isn't in the
logs, change your logging and next time it happens immediately zip up all the
log files and call me..."
Unfortunately with the defaults I find myself saying the above all the time.
Solr produces so much information that the logs roll over pretty quickly under
any kind of load, especially indexing. So the info I want is often gone anyway.
Which leads to another discussion about the default configuration ;)
I suspect this going to morph into a wider discussion about "logging in
general". It's grown ad-hoc so far, your comments about "is there a query log"
is a question I've heard from clients and on the user's list more than a few
times for instance.
I don't particularly care if the defaults are producing lots and lots and lots
of info, At present, though, we don't rationally group log messages, they're
mostly INFO. The important stuff is buried in a sea of messages. I can't grep
for WARN and be confident I'll see messages that matter for instance. Of course
"what matters" often is something arcane ;)...
If the consensus is to move a lot of our INFO messages to DEBUG and distribute
OOB with DEBUG enabled, that's fine as well.... I have about 4 different
versions of my Java program that I use to tease out different "stuff" from the
log file, mostly just to ignore noise. Which noise I need to filter out changes
with the problem unfortunately ;)
As an example of how logging has grown without any supervision... Did you know
that the "%C" bit in the layout pattern generates an exception for each and
every log line issued? Useful information to be sure when trying to analyze a
problem but costly steady-state. Not only do we generate a stack trace, but
also allocate memory, contribute to GC, etc.
That's an example of how we've viewed logging from a
development/troubleshooting PoV without considering the needs of the operations
folks. Much of this discussion will be about thinking about this from an
operations viewpoint too.
> Visit Solr logging, it's too noisy.
> -----------------------------------
>
> Key: SOLR-11934
> URL: https://issues.apache.org/jira/browse/SOLR-11934
> Project: Solr
> Issue Type: Improvement
> Security Level: Public(Default Security Level. Issues are Public)
> Reporter: Erick Erickson
> Assignee: Erick Erickson
> Priority: Major
>
> I think we have way too much INFO level logging. Or, perhaps more correctly,
> Solr logging needs to be examined and messages logged at an appropriate level.
> We log every update at an INFO level for instance. But I think we log LIR at
> INFO as well. As a sysadmin I don't care to have my logs polluted with a
> message for every update, but if I'm trying to keep my system healthy I want
> to see LIR messages and try to understand why.
> Plus, in large installations logging at INFO level is creating a _LOT_ of
> files.
> What I want to discuss on this JIRA is
> 1> What kinds of messages do we want log at WARN, INFO, DEBUG, and TRACE
> levels?
> 2> Who's the audience at each level? For a running system that's functioning,
> sysops folks would really like WARN messages that mean something need
> attention for instance. If I'm troubleshooting should I turn on INFO? DEBUG?
> TRACE?
> So let's say we get some kind of agreement as to the above. Then I propose
> three things
> 1> Someone (and probably me but all help gratefully accepted) needs to go
> through our logging and assign appropriate levels. This will take quite a
> while, I intend to work on it in small chunks.
> 2> Actually answer whether unnecessary objects are created when something
> like log.info("whatever {}", someObjectOrMethodCall); is invoked. Is this
> independent on the logging implementation used? The SLF4J and log4j seem a
> bit contradictory.
> 3> Maybe regularize log, logger, LOG as variable names, but that's a nit.
> As a tactical approach, I suggest we tag each LoggerFactory.getLogger in
> files we work on with //SOLR-(whatever number is assigned when I create
> this). We can remove them all later, but since I expect to approach this
> piecemeal it'd be nice to keep track of which files have been done already.
> Finally, I really really really don't want to do this all at once. There are
> 5-6 thousand log messages. Even at 1,000 a week that's 6 weeks, even starting
> now it would probably span the 7.3 release.
> This will probably be an umbrella issue so we can keep all the commits
> straight and people can volunteer to "fix the files in core" as a separate
> piece of work (hint).
> There are several existing JIRAs about logging in general, let's link them in
> here as well.
> Let the discussion begin!
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
