Hello,
We have been debugging the problems one of our UK Users has been seeing
where transfers of large (1 - 3GB) files into and out of our dpm SE
occasionally (but not all the time) fail at a random point mid-transfer.
The error message he sees are:
globus_xio: System error in write: Connection reset by peer
globus_xio: A system call failed: Connection reset by peer
He is using lcg-cp, with the srm2.2 protocol. The client tool versions are
lcg_util-1.6.17 and GFAL-client-1.10.18.
Has anyone seen this error message before, or something similar? And if
yes how did you debug the problem? I'm trying to understand the forensics
of it to track down the root this problem, the dpm experts have been
working with us and it doesn't look like a SRM server-side problem.
cheers,
Matt
|