Good day!
I've noticed that some of job that are landing on our resource center
(RRC-KI) are hanging at the stage of copying files to/from resource broker.
The process that is hanging is globus-url-copy. Attaching to that process
via strace revealed that the process does the following loop:
-----
select(9, [3], [], [], {0, 496906}) = 0 (Timeout)
gettimeofday({1148648859, 438672}, NULL) = 0
gettimeofday({1148648859, 438756}, NULL) = 0
gettimeofday({1148648859, 438840}, NULL) = 0
gettimeofday({1148648859, 438924}, NULL) = 0
select(9, [3], [], [], {0, 496957}) = 0 (Timeout)
gettimeofday({1148648859, 938629}, NULL) = 0
gettimeofday({1148648859, 938712}, NULL) = 0
gettimeofday({1148648859, 938795}, NULL) = 0
gettimeofday({1148648859, 938872}, NULL) = 0
select(9, [3], [], [], {0, 497009}) = 0 (Timeout)
gettimeofday({1148648860, 438583}, NULL) = 0
gettimeofday({1148648860, 438666}, NULL) = 0
gettimeofday({1148648860, 438750}, NULL) = 0
gettimeofday({1148648860, 438826}, NULL) = 0
select(9, [3], [], [], {0, 497055}) = 0 (Timeout)
gettimeofday({1148648860, 938542}, NULL) = 0
gettimeofday({1148648860, 938625}, NULL) = 0
gettimeofday({1148648860, 938711}, NULL) = 0
gettimeofday({1148648860, 938787}, NULL) = 0
select(9, [3], [], [], {0, 497094}) = 0 (Timeout)
gettimeofday({1148648861, 438544}, NULL) = 0
gettimeofday({1148648861, 438623}, NULL) = 0
gettimeofday({1148648861, 438717}, NULL) = 0
gettimeofday({1148648861, 438804}, NULL) = 0
-----
Has anyone seen such behaviour of globus-url-copy?
Thanks!
--
rea
|