[p4] Sanity check

[Robert Cowham]

You need to find the "base" file for anything missing/bad and then look to
see what has happened to that file on disk.

Note that because of the branching and "lazy copies" it is not always
obvious which underlying versioned file is bad/missing.

The brute force approach (and with older servers) p4v's revision graph is
great to sort this out. My little recipe, much exercised recently is to to
copy the path into clipboard, paste it into p4win to find the file, launch
rev graph and then click on the integrated tab and work my way back via the
intgrations tab and "sources contributing to this revision". It really is
quite fast. Then you do the obvious mapping between the earliest rev/file
and the depot map to find the source versioned file.

Note that with the latest version or two of the server, the above is even
more easily done with a single command (lurking in undoc):

C:\bruno_ws>p4 fstat -Oc //depot/jam/rel2.1/src/jam.c
... depotFile //depot/Jam/REL2.1/src/jam.c
... headAction branch
... headType text
... headTime 1106847349
... headRev 1
... headChange 749
... headModTime 1106847325
... lbrFile //depot/Jam/MAIN/src/jam.c		<==== what you want
... lbrRev 1.16
... lbrType text

Regards
Robert


> -----Original Message-----
> From: perforce-user-bounces at perforce.com 
> [mailto:perforce-user-bounces at perforce.com] On Behalf Of Smith, Jeff
> Sent: 28 February 2007 17:56
> To: perforce-user at perforce.com
> Subject: [p4] Sanity check
> 
> We thought we would perform a test restore of our Perforce 
> installation just to prove that we could.
>  
> We currently backup by performing a checkpoint and then 
> backing up that along with the depot.  The service is kept 
> live during this process.  As I understand it, the worst case 
> is that there will be revisions in the revision files that 
> would not have metadata and would therefore be effectively 
> lost.  Would this result in verify errors?
>  
> After restoring and reloading the database from the 
> checkpoint, there are a lot of MISSING! and BAD! verification 
> errors.  What is the most likely cause of these?
>  
> Thanks,
> Jeff