7 messages in com.perforce.perforce-user[p4] Re: p4 verify -q| From | Sent On | Attachments |
|---|---|---|
| Craig Allsop | 04 Feb 2004 15:57 | |
| ste...@vance.com | 04 Feb 2004 16:57 | |
| Craig Allsop | 05 Feb 2004 20:00 | |
| Rich Ireland | 05 Feb 2004 20:12 | |
| Robert Cowham | 06 Feb 2004 00:32 | |
| Craig Allsop | 06 Feb 2004 01:46 | |
| ste...@vance.com | 06 Feb 2004 08:21 |
| Subject: | [p4] Re: p4 verify -q |
|---|---|
| From: | Robert Cowham (rob...@vaccaperna.co.uk) |
| Date: | 02/06/2004 12:32:57 AM |
| List: | com.perforce.perforce-user |
A verify has to recalculate the digest and compare it with the stored one for every revision specified - otherwise it wouldn't detect files that have "gone bad" since verify was last done. Thus I wouldn't expect it to be faster. What is faster is the verify -u since it should be pretty much unnecessary (except for needing to be done once, and subsequently only for submits with old client versions).
The MD5 digest is used when doing things like diffs, e.g. if MD5s are different then do detailed diff, otherwise don't bother.
The backup server idea is a good one - tests your backups too.
FYI, this *should* be better with the latest server and client: http://www.perforce.com/perforce/doc.032/user/relnotes.txt
#47719 * ** The MD5 digest (fingerprint) of files are now computed and stored during submit. For best performance, use both a new 2003.2 client as well as server, as the new 2003.2 client programs offload the MD5 computation from the server. (Bug #2696).
Craig Allsop wrote:
Hi Steve.
We have dozens of file types :( You're suggestion of using the first revision only reduces it by an hour. :( There must be much bigger fish than us, do they not suffer this issue or not checkpoint daily? The server does not perfom well when doing a verify and having 30 people using it, it's a pity I can't have another computer performing the actual "verify". We have a seperate machine that rsyncs the files as a backup measure. Maybe I could configure it to unpack the checkpoint, create the db files, start a 2nd perforce server and run a verify on my backup instead. This would leave the live perforce server along except for a 20 minute nightly checkpoint.