DSXchange

Hi,

This Job run with Server (and not with PX ... Question of time)

I never see this message :
- : ds_ipcput - timeout waiting for mutex


the problem is :

We have a jobs using DataStage Link Partitioner.

Input to the link partitioner is a file (containing over 52,000,000 rows)
Each partition goes through several Shared Containers sequences.
There are 4 distinct Shared Containers.
Going through one of the partition,

- first Share Container will match the input record several hash files
- if record is unmatched, then it is be reported (second share container)
- If all is fine, then the third share container will get other informations
- Finally the forth share container post data on sequential file
and maintain key information in one hash file

These 4 Share Containers sequence are invoked 6 times using the Link Patitioning.

The problem is : after a while (it went even near the end of our 52,000,000 rows)
the job hangs and after we stop it provide log messages such as :

- : ds_ipcput - timeout waiting for mutex

Our DataStage job performance propertied are set to :
- Enable Row Buffer, Inter process
- Buffer size set to 128Kb
- Timeout set to 45 seconds

Pey

Pey,
This is a Windows message related to threads. I would not assume that the Timeout you set on the row buffering is related to the error message. I would contact Ascential if I were you. Do you have another server you could try it on?
John

Sorry, but I run on Unix. It's not a windows problem.

Pey

quote:Originally posted by Peytot
[br]Hi,

Our DataStage job performance propertied are set to :
- Enable Row Buffer, Inter process
- Buffer size set to 128Kb
- Timeout set to 45 seconds

Pey


Pey,

In my experience, it is usually caused by Inter Process Row Buffering and Enabling the pre-pload to memory option in hash files. You may experiment on these options:
1. disable all pre-load to memory settings, or
2. set the Timeout period to 0

regards,
vladimir