[cvsnt] rtag problem
arthur.barrett at march-hare.com
Tue Jun 19 11:13:32 BST 2007
I've never seen what you describe - if audit says the files were tagged then they were tagged...
The second tag will not have tagged files that already had the tag - hence why there were less tag records for the second tag - but does not explain why there were any tags at all the second time.
I'd need a trace of the error occuring to consider begin investigating.
From: cvsnt-bounces at cvsnt.org on behalf of Rick Martin
Sent: Tue 6/19/2007 3:43 AM
To: cvsnt at cvsnt.org
Subject: [cvsnt] rtag problem
Early today one of our developers ran an rtag to create a tag on a branch
for a module.
He subsequently ran an operation that relied on the new tag and the
I investigated and found that the tag was not on at least a fair number of
files in the module.
I had him re-run the process to create the tag and everything worked OK.
We have database auditing turned on and when I look at the tag log I can
see the original operation processed 5759 records. This is correct. The
files I spot checked that did not have the tag are including in the audit
database as having gotten the tag.
The audit DB shows that when he re-ran the rtag again it marked 4064 files.
Literally all of these files were in the first group, too. From what I can
tell the rtag operation does not create audit entries for the files where
the tag already existed (we did not have _F on).
The rtag call is programmatically generated so there isn't much chance of
the user entering the wrong values.
There are no other tag operations in the tag audit database for this time
frame until we re-ran the operation.
He did not capture the output from the original operation so I don't know
if something significant was returned but the audit DB certainly looks like
We're using 2.5.03 (Scorpio) Build 2382 for both client and server.
cvsnt mailing list
cvsnt at cvsnt.org
More information about the cvsnt