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

SubWCRev is terribly slow (on fresh working copies?!?)

From: Lübbe Onken <luebbe.tortoisesvn_at_gmail.com>
Date: Mon, 3 Aug 2015 10:42:35 +0200

Hi Folks,

I noticed the following strange behaviour of SubWCRev. I also found a
solution for the problem, which doesn't mean everything is fine, because
I'd like to know why this happens and if there is a way to prevent it from
happening in the first place.

Whenever I set up a new project on our Jenkins build server, the server
checks out a fresh working copy and builds the project in it. The size of a
fresh working copy is about 1.5 gigabytes, including build artifacts about
1.7 gigabytes.

Building a fresh project takes about one hour, which is incredibly slow.
Looking at the console output during the build I see that the build spend
several minutes (sic!) in each call to SubWCRev. SubWCRev is called many
times during the build, so these minutes really add up...

The solution is to run TortoiseSVN->Cleanup on the working copy. After that
a call to SubWCRev only takes a few seconds again and the build times are
back to the "normal" 15-17 minutes (factor four!)

The TortoiseSVN->Cleanup itself takes several minutes.
Jenkins checks out the working copies in svn 1.8 format using svnkit.

I reckon that TortoiseSVN->Cleanup repairs a somehow initially broken
working copy (rebuilds/creates indexes?!?)

Has anyone else encountered a similar problem?
What could be the reason for this behaviour?
What is your solution?

- Lübbe

Please help me get more space on Dropbox :)
  oo  // \\      "De Chelonian Mobile"
 (_,\/ \_/ \     TortoiseSVN
   \ \_/_\_/>    The coolest Interface to (Sub)Version Control
   /_/   \_\     http://tortoisesvn.net  PGP Key ID 0x23F511AB
To unsubscribe from this discussion, e-mail: [users-unsubscribe_at_tortoisesvn.tigris.org].
Received on 2015-08-03 10:42:44 CEST

This is an archived mail posted to the TortoiseSVN Users mailing list.