Testbed Support for GridPP member institutes
> [mailto:[log in to unmask]] On Behalf Of David Colling said:
> Alex claims there was
> no firewall at QMUL so does anybody have any ideas as to what could
have
> been blocking this?
Well, it could be an outbound firewall from RAL, or some other failure
of the gridftp server(s), but they seem less likely.
> Initially, it seemed to start off OK and an srm-get-metadata showed a
> file to exist, with zero bytes, but still I thought
> "Progress!" but then
> it failed with a timeout suggesting a network error. Thereafter, the
> subsequent re-tries ggave
Immediate guess is that the RAL dcache has died, e.g. I get:
lcg-cr --vo dteam file:///etc/group -d dcache.gridpp.rl.ac.uk
a system call failed (Connection refused)
advisoryDelete(User [name=dteam001, uid=36300, gid=24311,
root=/],pnfs/gridpp.rl.ac.uk/data/dteam/generated/2006-08-18/filee07e7ba
b-79b9-43de-ab5a-6840acc7632b) Error file does not exist, cannot delete
lcg_cr: Transport endpoint is not connected
On the other hand, for me qmul looks OK:
lcg-cr --vo dteam file:///etc/group -d se01.esc.qmul.ac.uk
guid:35589b7b-62fd-4819-8507-b1c1132a18c5
> What can be causing this? Could it be the different channels or what?
> How do I tell if it is the RAL end or the QMUL end or
> mismatch between the two.
As above - test the SEs individually. It's always better to test things
using the lowest-level tools you can and work up, e.g. bare
globus-url-copy (first and third party copies), edg-gridftp-ls etc,
lcg-utils, srmcp ... also try different combinations of sites, and if
possible different users/VOs/CAs/whatever.
> Never in all my years of HEP computing have I found anything so
> frustrating to pin down and debug.
You clearly haven't done much grid debugging over the last few years :)
> It really seems as though
> there are
> little packet gremlins sitting on the network connections or perhaps
> ghosts in the disk servers ... maybe I have been working too hard ;-)
But we all know that computers are totally predictable in their
behaviour ... (much like the weather :)
Stephen
|