Link Collector Problem

[cheerfu1]

Hi All,
I have encountered the following error in one of my datastage jobs using one Link Collector SOMETIMES which results in abort. After i reset it and run again, it would be ok. Thus, im puzzled as it would occur unpredictably as there are situations whereby continuous runs of 3-4 times are successful.

S10..MergeAll.Rejected: ds_ipcclose() - Error in mutex_unlock() - Operation not permitted.

DS10..MergeAll.Rejected: DSD.INTClose Error 11 in GCI close.

DS10..Compare: ds_ipcopen() - Error in open(/tmp/TMOPT.DS10..Compare) - A file or directory in the path name does not exist.

DS10..MergeAll: |Error 11 in GCI Link initialisation.|

Attempting to Cleanup after ABORT raised in stage DS10..MergeAll

DataStage Job 15 Phantom 8325
DS10..MergeAll.Rejected: DSD.INTClose Error 11 in GCI close.
DataStage Phantom Finished

Can anyone please enlighten on what may be the problem?

[chulett]

Just a guess, but I'd say you might be running out of temp space.

-craig

[dlvsrinivas]

Hi All,

Is there any solid reason for the above problem? I am also facing the same problem.

Regards,
Srinivas.

[PhilHibbs]

Hi All,
I have encountered the following error in one of my datastage jobs using one Link Collector SOMETIMES which results in abort. After i reset it and run again, it would be ok. Thus, im puzzled as it would occur unpredictably as there are situations whereby continuous runs of 3-4 times are successful.

I am having the same problem. The job failed twice, then worked. Here are some error messages:

Message: AR_CES_IM_J05_DMR3_DOC_Create_Segments..Collate_GL_entries.E1BPACGL09_vendor: ds_ipcflush() - Error in mutex_unlock() - Not owner
Message: AR_CES_IM_J05_DMR3_DOC_Create_Segments..Collate_GL_entries.E1BPACGL09_vendor: DSD.INTClose Error 11 in GCI close.
Message: AR_CES_IM_J05_DMR3_DOC_Create_Segments..Populate_customer_segments.E1BPACGL09_customer: Run stopped after 1547 rows

Any ideas? The only thing that I did before the third run (which worked) is to delete the destination files, but there was no problem with file permissions.

[chucksmith]

Some of your problems may be version specific. In version 7.something, the collector and partitioner were rewritten to perform non-blocking I/O. This solves most problems. Prior to the rewrite, if a collector or partitioner became starved for resources, it could hang. Uneven row counts on the partitioned links was also a problem at earlier releases.

I also suspect that having more processes trying to write to the same job log may also create a problem. However, this may be solved by changes in the uvconfig file, but I have not tested this, yet.

[chunsli]

Unix file permissions may cause this kind of problem if you run the job under different userIDs, expecially when the userIDs are in different groups and have different rights.

chunsli

Hi All,
I have encountered the following error in one of my datastage jobs using one Link Collector SOMETIMES which results in abort. After i reset it and run again, it would be ok. Thus, im puzzled as it would occur unpredictably as there are situations whereby continuous runs of 3-4 times are successful.

S10..MergeAll.Rejected: ds_ipcclose() - Error in mutex_unlock() - Operation not permitted.

DS10..MergeAll.Rejected: DSD.INTClose Error 11 in GCI close.

DS10..Compare: ds_ipcopen() - Error in open(/tmp/TMOPT.DS10..Compare) - A file or directory in the path name does not exist.

DS10..MergeAll: |Error 11 in GCI Link initialisation.|

Attempting to Cleanup after ABORT raised in stage DS10..MergeAll

DataStage Job 15 Phantom 8325
DS10..MergeAll.Rejected: DSD.INTClose Error 11 in GCI close.
DataStage Phantom Finished

Can anyone please enlighten on what may be the problem?

[changming]

I have same problem, so I seldom use link collector.
when I have to use, I use hash file. or I write into several sequentail file and then cat then after jobs.

[rumu]

Hi,
I think it is a performance issue,check out in performance tab which option is checked-user-project defaults if so then check all those parameter in administrator ,if not then check inter process under Enable row buffer option and increase the time out-say 60sec-since u r facing mutex error it might be problem of dead lock.

[PhilHibbs]

Unix file permissions may cause this kind of problem if you run the job under different userIDs,

Like I said, it was not a file permissions problem.

[tcj]

On a project I was working on we were having mutex errors while using link collectors. They were running datastage 7.5.1 on a AIX server. Seems that there was a glitch in he AIX operating system which was causing the mutex error on any jobs using a link collecter.

Anyone here that is having mutex errors running on a AIX server?

Tim