Page 1 of 1
STAGE has unexpected STATUS value
Posted: Thu May 26, 2005 1:33 am
by roy
Hi,
I have a DRS performing select max for a column to a Xfm poping a user status and outputing to /dev/null.
it was working fine till recently we started getting
the above message and the job aborts.
(searcing for the entire error got 4 results all leading to a forum page but no real post).
Thanks in advance,
Posted: Thu May 26, 2005 5:42 am
by chulett
Roy - not sure what you mean by "select max for a column to a Xfm poping a user status and outputing to /dev/null".

You might want to give more details on that.
Posted: Thu May 26, 2005 5:48 am
by Sainath.Srinivasan
I think he means writing the output to user status and connecting the link output to a (dummy) seq file with name as /dev/null in order to ignore any output from it.
Roy,
Did you check whether any value is populated into the user status?
Posted: Thu May 26, 2005 5:50 am
by chulett
Ok, I could see that. In addition to what was already asked, how exactly is the value getting 'popped' into the status area?
Posted: Thu May 26, 2005 7:24 am
by ray.wurlod
Something strange has been written, or failed to be written, to the RT_STATUSnn file record for that stage in the job. Clearing the status file for the job should remedy the problem.
Posted: Mon May 30, 2005 9:13 am
by roy
Thank for the replies All,
After some "changes"/recompiles and such it started working.
I'll try clearing the ststus file Ray next time

,
I still need to verify this being persistant and not a random fluke now and then.
(yes you got the idea Sainath.Srinivasan, I can't recall if it did

, if I get the answer to that I'll update this post)
Thaks Guys

,
Posted: Mon May 30, 2005 3:57 pm
by ray.wurlod
Recompiling, among many other things, does have the effect of resetting stage information in the status file. Or causing another wave number to be used next time the job is run, depending on internal conditions, which I'd prefer not to discuss.