Source Off Site quitting on large transfers

skrap1r0n · Post by **skrap1r0n** » Wed Jan 12, 2005 1:39 pm

Greetings,

We are having some issues with source off site timing out while doing a Get Latest Version. When I look at the services, it says it's started, however SoS isn't working. Everything works fine when I restart the service.

I have searched the Forum and located this thread. I followed some of the instructions to try and get a bit more info on how to fix this. Our SourceSafe DB is on the same machine and we are using a single DB, so the solution described in that thread does not apply.

I looked at the log.txt and I noticed that when it seems to fail, it tags the logfile with the following message:

1/12/2005 5:06:36 AM - Connection accepted from xxx.xxx.xxx.xxx:2363 on local address xxx.xxx.xxx.xxx:8081, session id is 9.
1/12/2005 5:06:41 AM - 9: Unable to get hostname from address: xxx.xxx.xxx.xxx
1/12/2005 5:25:38 AM - 9: Exception: Index was outside the bounds of the array.

The main issue we are having is that we have foreign developers working on web apps and there is a 13 hour time difference. SoS has canned out every day when they are doing a get. This is a new site with a lot of multimedia items in it and the site is about 860mb total.

Does anyone have a clue why SoS is just freezing up. Any help would be appreciated.

EDIT: We are using SoS v4.0.2 and VSS6.0d on a Windows server 2003 platform.

bfinney · Post by **bfinney** » Wed Jan 12, 2005 4:22 pm

Hmmm... this feels like the concurrency problem detailed here: http://support.sourcegear.com/viewtopic.php?t=1866 , but it's just a guess at this point.

Are you sure SOS is actually timing out? If so, the clients should receive a "server closed connection" message popup. If the client just hangs and never returns an error, it's very probably a concurrency problem.

Take a look at the article above and see if it may fit your situation. If not, we'll dig a little deeper.