Artifacts Resolving Failed, 80 Second Timeout

I'm getting a probjem with a build that has an artifact dependency. It was working but has stopped as the dependency zip file has grown in size - there seems to be an 80 second timeout that is being hit when downloading the artifact that gets hit and resets the connection. In the build log I get this message repeated 10 times as it retrys to download the artifact:

Resolving artifact dependencies (1m:20s)
[21:22:09][Resolving artifact dependencies] Destination directory cleaned: D:\TeamCity\buildAgent\     
[21:23:29][Resolving artifact dependencies] Failed to resolve artifact dependency <??? :: ???, ??? [???]>: Recoverable failure while downloading artifact ???.zip: java.net.SocketException: Connection reset (jetbrains.buildServer.artifacts.impl.RecoverableResolvingException)
[21:23:29][Resolving artifact dependencies] 1 recoverable failures while resolving 1 artifact dependencies
[21:23:29]1 recoverable failures while resolving 1 artifact dependencies



If the artifact can be downloaded in under 80 seconds then it will work every time, every time it hits 80 then the connection is reset.

Does anyone have any ideas of where this timeout comes from and how I can increase it?
4 comments
Comment actions Permalink

Hello Ian,

TeamCity does not have time limits in artifact resolution process.

What is your network configuration? Do you have any proxies/reverse proxies or firewalls?

0
Comment actions Permalink

Hi Nikita,

Thanks, that's useful to know. Our TeamCity server is setup on a windows Amazon EC2 instance, in this case the Agent is actually running on the same server with no proxy.

Since it's unlikely a TeamCity problem I'll start to look elsewhere and update when I get to the bottom of it.

0
Comment actions Permalink

in the meantime, could you please attach relevant parts of teamcity-server.log and teamcity-agent.log to this forum thread?

0
Comment actions Permalink

It turns out that this was something broken on the EC2 instance, for some reason it's network transfer had gone very slow. Killing it and sparking up a new one has resolved my problem.

0

Please sign in to leave a comment.