On Thu, Mar 17, 2005 at 03:38:17PM -0500, Chris Penney wrote:
> > > 5) Are there any tuning options I should consider?
> > 
> > If you're typical workload is very meta-data intensive with lots of
> > threads etc. you might want to increase the nTxLock or nTxBlock module
> > parameters, but in general it shouldn't be necessary.
> 
> This box will be very write oriented (80% writes) and will service
> proablly 30-40 batch jobs running on linux clusters.  It's very
> bursty, but not monumental.  There are a few applications that do a
> lot of small file i/o (lots of seeks, writes, commits).  The round the
> clock data rate is probablly around 15MB/s.
> 
>    Chris

Ok, if you do testing with lots of threads and
create/appends/deletes/etc look at   /proc/fs/jfs/txstats 

It'll look like this:

JFS TxStats
===========
calls to txBegin = 0
txBegin blocked by sync barrier = 0
txBegin blocked by tlocks low = 0
txBegin blocked by no free tid = 0
calls to txBeginAnon = 0
txBeginAnon blocked by sync barrier = 0
txBeginAnon blocked by tlocks low = 0
calls to txLockAlloc = 0
tLockAlloc blocked by no free lock = 0


If you see non-zero numbers for "txBeginAnon blocked by tlocks low "
or "tLockAlloc blocked by no free lock" then you might benefit from
more txlocks and txblocks.

Sonny


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
Jfs-discussion mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/jfs-discussion

Reply via email to