Post questions here relative to DataStage Server Edition for such areas as Server job design, DS Basic, Routines, Job Sequences, etc.
Moderators: chulett , rschirm , roy
dspxguy
Participant
Posts: 156 Joined: Thu May 24, 2007 4:09 pm
Location: Simi Valley, CA
Post
by dspxguy » Mon Apr 21, 2008 1:22 pm
hi, we got this error message in a server job
Code: Select all
chekstglink: ds_ipcopen() - Error in open(/tmp/ade.projecta.jobname.RecodeUnitCnts_tran-Combine.chekstglink) - No such file or directory
later on the job finished successfully. I checked the /tmp space and it was not full and there were no permission issues as well.
what was wrong?
appreciate your input.
Thanks
kumar_s
Charter Member
Posts: 5245 Joined: Thu Jun 16, 2005 11:00 pm
Post
by kumar_s » Mon Apr 21, 2008 1:24 pm
What is the job design? Are you using any IPC stage or Link collector etc?
Impossible doesn't mean 'it is not possible' actually means... 'NOBODY HAS DONE IT SO FAR'
dspxguy
Participant
Posts: 156 Joined: Thu May 24, 2007 4:09 pm
Location: Simi Valley, CA
Post
by dspxguy » Mon Apr 21, 2008 1:56 pm
Code: Select all
SeqFile --- Hash
|
xfm -------->xfm ------>xfm------->LinkPartitioner
| | |
| Hash Seq File
DB2 -------->Linkcoltr
|
DB2 -------->xfm
using link collector and link partitioner stages,
kumar_s
Charter Member
Posts: 5245 Joined: Thu Jun 16, 2005 11:00 pm
Post
by kumar_s » Mon Apr 21, 2008 2:06 pm
Have you selected "Enable Row Buffer" at job property level to Inter Process?
Impossible doesn't mean 'it is not possible' actually means... 'NOBODY HAS DONE IT SO FAR'
dspxguy
Participant
Posts: 156 Joined: Thu May 24, 2007 4:09 pm
Location: Simi Valley, CA
Post
by dspxguy » Mon Apr 21, 2008 3:20 pm
kumar_s wrote: Have you selected "Enable Row Buffer" at job property level to Inter Process? ...
yes
ray.wurlod
Participant
Posts: 54607 Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:
Post
by ray.wurlod » Mon Apr 21, 2008 4:47 pm
Checking temp space "later" is of no value. You have to check it "during".
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
dspxguy
Participant
Posts: 156 Joined: Thu May 24, 2007 4:09 pm
Location: Simi Valley, CA
Post
by dspxguy » Mon Apr 21, 2008 5:02 pm
Ray,
i totally agree with that statement.
i am just curious why did it run fine after the abort., maybe 1GB tmp space is not enough for the jobs..
ray.wurlod
Participant
Posts: 54607 Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:
Post
by ray.wurlod » Mon Apr 21, 2008 5:15 pm
It may not be, as you say. For example if you are processing (sorting, maybe) 10GB of data, then 1GB of scratch space will possibly not be enough: definitely won't be enough if you only have 4GB of memory in the machine.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
dspxguy
Participant
Posts: 156 Joined: Thu May 24, 2007 4:09 pm
Location: Simi Valley, CA
Post
by dspxguy » Wed Apr 30, 2008 9:44 am
hi, thanks for the responses. another question,
what is the purpose of defining the varaiable TMPDIR in adminsitrator when we already use UVTEMP in uvconfig?
In some of our environments we have defined both and in some we are only using UVTEMP space.
is TMPDIR a backup space, like when scratch will be full then it will be used?
Thanks
ray.wurlod
Participant
Posts: 54607 Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:
Post
by ray.wurlod » Wed Apr 30, 2008 2:51 pm
Yes, TMPDIR is the third option for scratch space in parallel jobs. UVTEMP is scratch space used by server jobs. Custom stages may use either, at the whim of their author.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
dspxguy
Participant
Posts: 156 Joined: Thu May 24, 2007 4:09 pm
Location: Simi Valley, CA
Post
by dspxguy » Thu May 01, 2008 3:08 pm
Thanks for the info,