[p4] What actually happens during checkpoint

Jamie.Echlin at barclayscapital.com Jamie.Echlin at barclayscapital.com
Fri Mar 6 08:25:39 PST 2009


Thanks for all the input I've had on this question, if you replied to me
off-list sorry if I didn't manage to reply, but all much appreciated.

So, I know the checkpoint speed is related to the amount of metadata and
the fragmentation of the btrees, ie how long it's been and how much
activity has happened since the last rebuild. I'm reducing the metadata
by binning old clients and users, trying to enforce more but tighter
views, preventing the continuous integration tools from doing a full
labelsync, and use automatic labels instead (labels table is 30G alone).

After a rebuild of the database I got back about 25G, and subsequent
checkpoint operations seem to be faster - I'm not going to say exactly
how much faster because I don't exactly have a quiescent server to test
this on.

Unfortunately I can't test this on local disk, it's just not an option
to even test it without first buying our own 15 terabyte EMC array...
Don't ask. Also not enough space to test with an uncompressed
checkpoint.

It does seem that XFS is faster than ext3, and ext3 is faster when
mounted with data=writeback, so I'll be looking at both of those
options. Support are going to send me a filesystem performance analysis
tool so I'll take it from there.

As I said we're underpowered in the memory department, an upgrade is on
its way.

In fact, I've attached the complete list of tasks to tackle this seeing
as everyone's been so helpful ;-)

Thanks again,
jamie

> -----Original Message-----
> From: Smith, Jeff [mailto:jsmith at medplus.com] 
> Sent: 05 March 2009 18:26
> To: Echlin, Jamie: IT (LDN); geoffrey.mantel at sap.com; 
> perforce-user at perforce.com
> Subject: RE: [p4] What actually happens during checkpoint
> 
> If there is anyway that you can do a local test to take the 
> SAN out of the equation it would be helpful.  Set up a second 

_______________________________________________

This e-mail may contain information that is confidential, privileged or otherwise protected from disclosure. If you are not an intended recipient of this e-mail, do not duplicate or redistribute it by any means. Please delete it and any attachments and notify the sender that you have received it in error. Unless specifically indicated, this e-mail is not an offer to buy or sell or a solicitation to buy or sell any securities, investment products or other financial product or service, an official confirmation of any transaction, or an official statement of Barclays. Any views or opinions presented are solely those of the author and do not necessarily represent those of Barclays. This e-mail is subject to terms available at the following link: www.barcap.com/emaildisclaimer. By messaging with Barclays you consent to the foregoing.  Barclays Capital is the investment banking division of Barclays Bank PLC, a company registered in England (number 1026167) with its registered office at 1 Churchill Place, London, E14 5HP.  This email may relate to or be sent from other members of the Barclays Group.
_______________________________________________


More information about the perforce-user mailing list