[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: Required version of APR

From: Stefan Sperling <stsp_at_elego.de>
Date: Fri, 7 Feb 2014 19:55:15 +0100

On Fri, Feb 07, 2014 at 04:43:32PM +0000, Philip Cornelius wrote:
> Thanks for your response..
> > What do you mean? A revision doesn't exist before it is committed.
> You wrote:
> > > Using a binary file that had been committed to a revision that was
> > > corrupt
> That sounded like a pre-condition for the bug. 'had been' i.e. were you
> able to reproduce this on a brand new known good repository. Or was the
> repository already in some kind of corrupt state.

A file F.msi was committed to a production repository R1,
in a revision which svnadmin verify then flagged as corrupt.
There was no visible error during the commit.
I wasn't around when this happened.

From the working copy this commit had been made from, we obtained a
good copy of the original F.msi.

We were able to occasionally reproduce the corruption on an unrelated
repository R2 by committing F.msi over and over again. About one in
one hundred commits of F.msi to R2 caused corrupt revisions.

> I am with you about getting off the old APR.. but I want to understand if
> this is urgent.. or could perhaps wait 6 months as we plan a move to a
> newer RHEL.

I think it is urgent. But if you don't have corrupt revisions already,
you might never see any. I cannot give better advice since, as already
explained, the exact details are unknown. I can't even tell if other
people will see the same 1/100 ratio or not.
Received on 2014-02-07 19:55:54 CET

This is an archived mail posted to the Subversion Dev mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.