pcp
[Top] [All Lists]

Re: [pcp] Oracle connection debugging (was Re: Handling Oracle PMDA Late

To: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>, Nathan Scott <nathans@xxxxxxxxxx>
Subject: Re: [pcp] Oracle connection debugging (was Re: Handling Oracle PMDA Latencies)
From: Marko Myllynen <myllynen@xxxxxxxxxx>
Date: Thu, 19 May 2016 16:02:44 +0300
Cc: pcp developers <pcp@xxxxxxxxxxx>
Delivered-to: pcp@xxxxxxxxxxx
In-reply-to: <573DB415.7000807@xxxxxxxxxxxxxxxx>
Organization: Red Hat
References: <56F25541.9020602@xxxxxxxxxx> <570F511E.5000605@xxxxxxxxxx> <1512930308.40394593.1460673441009.JavaMail.zimbra@xxxxxxxxxx> <57108708.3080906@xxxxxxxxxx> <571092DF.8050409@xxxxxxxxxx> <57175FC8.2000600@xxxxxxxxxx> <1558022602.42320984.1461208897951.JavaMail.zimbra@xxxxxxxxxx> <57395F04.2090909@xxxxxxxxxx> <1695396289.47966126.1463381940778.JavaMail.zimbra@xxxxxxxxxx> <573D897A.5070804@xxxxxxxxxx> <573DB415.7000807@xxxxxxxxxxxxxxxx>
Reply-to: Marko Myllynen <myllynen@xxxxxxxxxx>
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.8.0
Hi,

On 2016-05-19 15:39, Ken McDonell wrote:
> On 19/05/16 19:38, Marko Myllynen wrote:
>> ...
> 
> Just a reminder, two-threaded PMDAs (one for PMCD, one for SQL) and
> PM_ERR_PMDANOTREADY were invented in the first place to accommodate slow
> DBMS PMDAs.

Thanks for pointing that out - Nathan, does this sound at all feasible
/helpful for the Oracle PMDA?

FWIW, I checked how long the Oracle AWR report generation takes, it's
around 10 seconds which the local DBAs consider "normal." So perhaps we
can't expect the same approach which works with local /proc files to
work as-is with a massive Oracle instance under extreme load but rather
come up with something else, perhaps like Ken suggested above?

Thanks,

-- 
Marko Myllynen

<Prev in Thread] Current Thread [Next in Thread>