sks-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Sks-devel] Min. Requirement for SKS Version in the Pool


From: John Clizbe
Subject: Re: [Sks-devel] Min. Requirement for SKS Version in the Pool
Date: Mon, 25 Jun 2012 01:39:07 -0500
User-agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.1.20pre) Gecko/20110606 Mnenhy/0.8.5 SeaMonkey/2.0.15pre

Daniel Kahn Gillmor wrote:
> On 06/25/2012 02:16 AM, John Clizbe wrote:
>> After Christoph's last email re mutex_set_max I checked my own databases, 
>> both
>> of which were set to 64K.  PTree was almost equally split between in-use and
>> free.  KDB was very close to running out with only about 3K left of the 64K.
> 
> Would you mind showing how you did this check?  What architecture was
> the machine you checked?  I'd be happy to corroborate.
> 
>> With the larger number of keys now in the DB, perhaps a better value for
>> KDB/DB_CONFIG is
>> 
>>     mutex_set_max    98304
> 
> Can this change be made live in the DB_CONFIG file, while sks is
> operating?  or should we shut down sks, make the change, and then restart?

DB_CONFIG is read when the database is opened. You can make the edit, then do
a shutdown-start. I do a shutdown, db_recover -h KDB, sks start

> Is there some way that this can be auto-tuned?  What are the
> consequences of setting the value even higher?

The mutex region grows from ~7MB to ~12MB. BDB sets an initial mutex count of
57435. It when then grow the mutex pool as it needs more. mutex_set_max is
just an upper bound on that growth. You can see the mutex usage with

    db_stat -xh DB

"Normally," BDB's defaults are adequate. Tuning is more of an
application-specific "black art".

> Thanks for helping me get my head around bdb configuration!

Thank Jeff Johnson for helping me.

-John



reply via email to

[Prev in Thread] Current Thread [Next in Thread]