ERROR CODE -14 ( PATCH)

myukassign · Post by **myukassign** » Mon Jun 07, 2010 1:05 am

Hi,

I did my homework by searching the dschange for code -14 error code. I realised the very common reason for this error is.

A. Server overload.

B. If a number of jobs triggering at the same time, and one of those jobs is not responding with in 60 seconds of turnaroud time.

I saw the solutons as to reduce the no of jobs triggering at the same time. / Check the server overload etc.

I also found an interesting solution to install for IBM path for AIX to increase the job timeout period.

The help I need is:

1. Please tell me that patch number? or whats the referece number of that patch to request for it.

2. Once I install the path, is there any environment variables needs to set to increase the timeout period.

Many many thanks in advance.

Arun

chulett · Post by **chulett** » Mon Jun 07, 2010 6:11 am

Never found any need for a patch, this is easy enough to never trigger by proper job management. And by 'easy enough', I mean I never had to go out of my way or worry about this in any way in spite of typically needing to run many (many) hundreds of jobs daily.

YMMV of course, but that's been my experience.

myukassign · Post by **myukassign** » Mon Jun 07, 2010 6:15 am

chulett wrote:Never found any need for a patch, this is easy enough to never trigger by proper job management. And by 'easy enough', I mean I never had to go out of my way or worry about this in any way in spite of typically needing to run many (many) hundreds of jobs daily.

YMMV of course, but that's been my experience.

Craig.. Please tell me what exactly I need to do.... I am not seeing my Sequncer is triggering so many jobs at a time....

Since this is a shared server and it may be possible that so many other projects ppl jobs may triggering at the same time which I have not much control ....

Please help.... Thanks for helping me in advance...

chulett · Post by **chulett** » Mon Jun 07, 2010 6:36 am

There is no 'exact' and not much to help to be given other than 'run fewer jobs at the same time'. Typically this means start rather than run as it is a startup issue.

On a 'shared server' you'd have to do this at an enterprise level, working with the other teams to minimize the problem. How busy is the DataStage server when you see this? How many jobs is your 'not so many' Sequence job controlling?

myukassign · Post by **myukassign** » Mon Jun 07, 2010 6:44 am

chulett wrote:There is no 'exact' and not much to help to be given other than 'run fewer jobs at the same time'. Typically this means start rather than run as it is a startup issue.

On a 'shared server' you'd have to do this at an enterprise level, working with the other teams to minimize the problem. How busy is the DataStage server when you see this? How many jobs is your 'not so many' Sequence job controlling?

oho.. hmmm ... I understood...

How to get the number of jobs running at my server at a point of time....

and how to identify how busy is my server ....

chulett · Post by **chulett** » Mon Jun 07, 2010 7:06 am

Grep for processes, something along the lines of:

Code: Select all

ps -ef |grep phantom | grep -v grep

Should get you started. As to the general 'busy' question, work with your SysAdmin for that.

myukassign · Post by **myukassign** » Mon Jun 07, 2010 7:37 am

chulett wrote:Grep for processes, something along the lines of:
Code: Select all
ps -ef |grep phantom | grep -v grep
Should get you started. As to the general 'busy' question, work with your SysAdmin for that.

One last clarification Craig....

How many jobs I can trigger simultaneously in a sequencer... It must be dependent on number of scenarios like u said...processor..memory resources etc...

But can you suggest an ideal approximate number which is potentially low risk.

Dont't give me as 1

chulett · Post by **chulett** » Mon Jun 07, 2010 8:02 am

How about... more than one?

There's no way for anyone to answer that, too many variables in both the hardware and exactly what your Sequence and the jobs it runs do for a cogent answer. You have to do something like they did at Los Alamos when they tickled the dragon's tail - ramp things up until they almost go critical and then back it back down a bit.

myukassign · Post by **myukassign** » Mon Jun 07, 2010 8:09 am

chulett wrote:How about... more than one?

There's no way for anyone to answer that, too many variables in both the hardware and exactly what your Sequence and the jobs it runs do for a cogent answer. You have to do something like they did at Los Alamos when they tickled the dragon's tail - ramp things up until they almost go critical and then back it back down a bit.

Nice example...I have my answers in place.....

Thanks a lot Craig... Let me play with this dragon tail