View previous topic :: View next topic
|
Author |
Message |
vasanthz
Global Moderator
Joined: 28 Aug 2007 Posts: 1742 Location: Tirupur, India
|
|
|
|
Hi,
Our system jobs rely on tape datasets for input or output.
Sometimes our jobs go on a long wait & waits for the tape to be mounted.
There was once a situation in which our jobs were in long wait for 2 or 3 hours. When I contacted the operations, they informed that there was a rogue job run by a user(neighbouring team) which occupied all the tape drives and still executing and this caused the long wait in all other jobs using tapes.
Could you please let me know how to determine the number of tape drives available and how find out if a job is using more tape drives?
I asked the ops guys at my place and they say that they find it using console. But I dont have access to console
Is there any command or facility to view this something like TSO ISRDDN E for viewing contentions.
We have CA-1 TMS at our shop.
Thanks, |
|
Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006 Posts: 19244 Location: Inside the Matrix
|
|
|
|
Hi Vasanthz,
I believe this is more a mangement problem than a technical issue.
If there are no "rules" in place on the system as to how many tapes drives a job may allocate, i believe you are stuck.
If there are rules and the "rogue job" violates the rules, it can (and probably should) be canceled. |
|
Back to top |
|
|
vasanthz
Global Moderator
Joined: 28 Aug 2007 Posts: 1742 Location: Tirupur, India
|
|
|
|
Hi D,
Thanks for your view, I agree that there has to be a rule for the max number of tape drives a user could use.
We have Omegamon and it had a screen for monitoring tape utilization.
omegamon -> tape -> option 1(details)
Code: |
+----------------------------------------------------------------------------+
¦ DevNo ¦ Volume ¦ User ¦ Status ¦ I/O's ¦Prm Err¦Tmp Err¦
¦--------+--------+----------+---------------------+---------+-------+-------¦
¦ ABCD ¦ NONE ¦ ¦ FREE ¦ 131,914 ¦ ¦ ¦
¦ B019 ¦ NONE ¦ ¦ FREE ¦ 581,688 ¦ ¦ ¦
¦ BCDE ¦ NONE ¦ ¦ FREE ¦ 637,047 ¦ ¦ ¦
¦ BDFE ¦ NONE ¦ ¦ FREE ¦ 667,981 ¦ ¦ ¦
¦ BK34 ¦ A12345 ¦ XXXXXXXX ¦ ALLOC ¦ 196,873 ¦ ¦ ¦
¦ SD34 ¦ ¦ YYYYYYYY ¦ ALLOC ¦ 018,059 ¦ ¦ ¦
¦ ERTF ¦ A33456 ¦ ZZZZZZZZ ¦ ALLOC ¦ 716,751 ¦ ¦ ¦ |
Regards, |
|
Back to top |
|
|
PeterHolland
Global Moderator
Joined: 27 Oct 2009 Posts: 2481 Location: Netherlands, Amstelveen
|
|
|
|
So now you are going to look at the screen, and refresh it every 2 seconds to see if there are free devices, and then submit your job? |
|
Back to top |
|
|
vasanthz
Global Moderator
Joined: 28 Aug 2007 Posts: 1742 Location: Tirupur, India
|
|
|
|
No, the production jobs would run as scheduled, its diagnostic tool if there are jobs waiting for a long time for tape mount.
Also have to talk to tape operations if they could restrict the number of drives used by a user. Its upto them.
Regards, |
|
Back to top |
|
|
Anuj Dhawan
Superior Member
Joined: 22 Apr 2006 Posts: 6250 Location: Mumbai, India
|
|
|
|
Vasanth - Possibly you can experiment with 1. UNIT=AFF=ddname and/or 2. "ping pong" tape drives.
To "ping pong" tape drives for concatenated input file(s) on 'n' physical drives, just specify UNIT=(TAPE,n) on the first DD statement.
eg. by using UNIT=(TAPE,2)
If the dataset is a multi-volume one, it will mount the first volume to 1 tape drive and the second volume to another tape drive, so when the job has finished with the first volume, it will immediately start processing the second volume and rewind the first volume and then mount the third volume whilst the second volume is being processed and so on. You can also use this for your output DD. |
|
Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006 Posts: 19244 Location: Inside the Matrix
|
|
|
|
Hello,
Unfortunately, neither will help if there are zero drives available to allocate (which is what i understood to be the problem). |
|
Back to top |
|
|
|