[Swift-user] Swift running errors

lixi at uchicago.edu lixi at uchicago.edu
Wed Feb 20 09:58:56 CST 2008


>> Failed to transfer kickstart records from workflowtest-
>> 20080219-1447-1hztqje9/kickstart/8/CIT_CMS_T2Exception in 
>> getFile

>Sometimes this happens because there was an error running 
you job for some 
>other reason (so the job didn't run, a kickstart record 
wasn't generated 
>and it couldn't be transfered). Have a look in the log file 
for an error 
>earlier than that (or put the log files online some so I 
can look).

This log file is in terminable.ci.uchicago.edu:
/home/lixi/swift/test/newtest/1000nodes/remote/workflowtest-
20080219-1447-1hztqje9.log

>> 2. When runing a workflow which involves 1000nodes, I 
>> encounter the following errors very frequently, but not 
all 
>> the time:
[..]
>>         java.io.IOException: Too many open files

>What machine are you running on? Have you changed any 
configuration 
>parameters? A basic 1000 job workflow should run fine with 
the default 
>settings.

>On the machine you are running on, type   ulimit -a   and 
paste that here.

Yesterday I was running that on login.ci.uchicago.edu, but 
it crashed last night, so I tried ulimit -a on 
terminable.ci.uchicago.edu and got:
[lixi at terminable ~]$ ulimit -a
core file size          (blocks, -c) 0
data seg size           (kbytes, -d) unlimited
file size               (blocks, -f) unlimited
pending signals                 (-i) 1024
max locked memory       (kbytes, -l) 32
max memory size         (kbytes, -m) unlimited
open files                      (-n) 1024
pipe size            (512 bytes, -p) 8
POSIX message queues     (bytes, -q) 819200
stack size              (kbytes, -s) 10240
cpu time               (seconds, -t) unlimited
max user processes              (-u) 15871
virtual memory          (kbytes, -v) unlimited

Does it mean that I can only at most open 1024 files at one 
time?

Thanks,

Xi



More information about the Swift-user mailing list