Unable Create Copy or Save As Job

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

Post Reply
ray.wurlod
Participant
Posts: 54607
Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:

Post by ray.wurlod »

Search would be a good idea.

There are very few reasons why an operating system file (in this case RT_CONFIG20857) can not be created. In summary they are (a) permissions, (b) disk full, (c) disk quota exceeded or (d) limit on number of subdirectories in a directory exceeded.

Do you really have more than 20,000 jobs in one project?
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
my_stm
Premium Member
Premium Member
Posts: 58
Joined: Mon Mar 19, 2007 9:49 pm
Location: MY

Post by my_stm »

ray.wurlod wrote:Search would be a good idea.

There are very few reasons why an operating system file (in this case RT_CONFIG20857) can not be created. In summary they are (a) permissions, (b) disk full, (c) disk quota exceeded or (d) limit on number of subdirectories in a directory exceeded.

Do you really have more than 20,000 jobs in one project?
Thank you for your replied. we are now doing research on this issues. :roll:
PhilHibbs
Premium Member
Premium Member
Posts: 1044
Joined: Wed Sep 29, 2004 3:30 am
Location: Nottingham, UK
Contact:

Post by PhilHibbs »

I am also having this problem, when it occurred the project directory contained 32766 subdirectories. Is there any way to lift this limit on AIX?
Phil Hibbs | Capgemini
Technical Consultant
ArndW
Participant
Posts: 16318
Joined: Tue Nov 16, 2004 9:08 am
Location: Germany
Contact:

Post by ArndW »

The 32766 limit is UNIX one; the inode/entry table for a directory is limited to this number of files. When this happened at a customer site with lots of jobs, I did a quickie fix of resizing all the RT_STATUSnnn files from dynamic to static files. The better solution is to "calve" a new project when your old one reaches this size.
PhilHibbs
Premium Member
Premium Member
Posts: 1044
Joined: Wed Sep 29, 2004 3:30 am
Location: Nottingham, UK
Contact:

Post by PhilHibbs »

ArndW wrote:The better solution is to "calve" a new project when your old one reaches this size.
We have a ton of old jobs that we can delete. I'll have to write a little script that checks the directory count in each project so I can see this coming up. Something that does this in each one...

Code: Select all

ls -l|grep ^d|wc -l
Or, I could just look at the parent directory, that's what the number after the flags is isn't it?
Phil Hibbs | Capgemini
Technical Consultant
chulett
Charter Member
Charter Member
Posts: 43085
Joined: Tue Nov 12, 2002 4:34 pm
Location: Denver, CO

Post by chulett »

The number after the flags? :?
-craig

"You can never have too many knives" -- Logan Nine Fingers
PhilHibbs
Premium Member
Premium Member
Posts: 1044
Joined: Wed Sep 29, 2004 3:30 am
Location: Nottingham, UK
Contact:

Post by PhilHibbs »

Code: Select all

phibbs@tsaix60 # ls -l
total 24840
drwxrwxr-x 24257 dsadm    dstage      3239936 28 Jul 14:31 C1PDM_Dev
drwxrwxr-x 20238 dsadm    dstage      1613824 18 Jul 02:48 C1PDM_Prod
drwxrwxr-x 7259 dsadm    dstage       643072 07 Apr 13:24 C1PDM_QA
drwxrwxr-x 8459 dsadm    dstage       557056 27 May 10:01 C1PDM_QCP4400
drwxrwxr-x 15296 dsadm    dstage      1695744 14 Jul 11:51 C1PDM_QCP4401
drwxrwxr-x 5778 dsadm    dstage       483328 16 Jul 06:13 C1PDM_QCP4402
drwxrwxr-x  107 dsadm    dstage         8192 18 Jul 02:48 C1PDM_QDV8231
drwxrwxr-x 17988 dsadm    dstage      1638400 28 Jul 14:05 C1PDM_QTS4331
drwxrwxr-x  263 dsadm    dstage        16384 09 Jul 09:50 C1PDM_QTS8331
drwxrwxr-x 1325 dsadm    dstage        73728 09 Jul 09:50 C1PDM_QTS8351
drwxrwxr-x 1595 dsadm    dstage       114688 28 Jul 13:16 C1PDM_QTS8352
drwxrwxr-x 17119 dsadm    dstage      1540096 27 Jul 17:30 C1PDM_QXR4331
drwxrwxr-x 5849 dsadm    dstage       462848 28 Jul 11:09 C1PDM_QXR4352
drwxrwxr-x 6443 dsadm    dstage       446464 27 Jul 21:12 C1PDM_Scratch
drwxrwxr-x   47 dsadm    dstage         4096 01 Jun 16:19 test
phibbs@tsaix60 #
Phil Hibbs | Capgemini
Technical Consultant
chulett
Charter Member
Charter Member
Posts: 43085
Joined: Tue Nov 12, 2002 4:34 pm
Location: Denver, CO

Post by chulett »

So, for example in the first line the number 24257? I'm not sure off the top of my head what it is but the man page should let you know.
-craig

"You can never have too many knives" -- Logan Nine Fingers
PhilHibbs
Premium Member
Premium Member
Posts: 1044
Joined: Wed Sep 29, 2004 3:30 am
Location: Nottingham, UK
Contact:

Post by PhilHibbs »

DataStage appears to create 6 directories for each job in a project. This limits a project to slightly over 5400 jobs. Does that seem reasonable? Well, we're going to blow that limit by a significant margin.

I'm going to have to create a new production project, which is kind of annoying as I have a lot of scripts that will break, and will be harder to use when I fix them as I'll have to pass an extra parameter to specify which project the jobs are in.

I'm also a little disappointed that DataStage isn't aware of this limit, and didn't put up a nice "Your project directory is full, please delete any un-necessary jobs or create a new project" message.
Phil Hibbs | Capgemini
Technical Consultant
ray.wurlod
Participant
Posts: 54607
Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:

Post by ray.wurlod »

PhilHibbs wrote:DataStage appears to create 6 directories for each job in a project.
For server jobs they are (where nnn is the job number):
DS_TEMPnnn
RT_BPnnn
RT_BPnnn.O
RT_CONFIGnnn
RT_LOGnnn
RT_STATUSnnn

Parallel jobs also create RT_SCnnn
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
ArndW
Participant
Posts: 16318
Joined: Tue Nov 16, 2004 9:08 am
Location: Germany
Contact:

Post by ArndW »

You can safely do a RESIZE of the DS_TEMPnnn, RT_CONFIGnnn and RT_STATUSnnn to static hashed files with a modulo of 1, this saves 3 subdirectories and 9 inodes per job.
PhilHibbs
Premium Member
Premium Member
Posts: 1044
Joined: Wed Sep 29, 2004 3:30 am
Location: Nottingham, UK
Contact:

Post by PhilHibbs »

Interesting. I mentioned this to my boss but he ruled it out as an option, we're going to create a new project. Oh well.
Phil Hibbs | Capgemini
Technical Consultant
Post Reply