how to fix archiving job errors

[Former Member]

Hello Experts,

We have daily batch runs for Archiving , the Basis team forced killed some of these jobs due to performance issues.

the situtation now is that there were also some spawn jobs (store, delete) jobs which were also cancelled (ripple affect).

now when I start the delete job manually for one of the objects RV_LIKP, I get the following error.

Any suggestions on how to resolve this issue?

Job log overview for job:    ARV_RV_LIKP_DEL20140311053442 / 05344200

Date       Time     Message text                                                                     Message class Message no. Message type

03/11/2014 05:34:42 Job started                                                                           00           516          S
03/11/2014 05:34:42 Step 001 started (program S3LIKPDLS, variant Z_ARCHIVE_PROD, user ID ARCH-BATCH)      00           550          S
03/11/2014 05:34:42 Archive file 013513-001RV_LIKP is being verified                                      BA           238          S
03/11/2014 05:34:43 Archive file 013513-001RV_LIKP contains errors from offset 25,504,229                 BA           192          S
03/11/2014 05:34:43 1 archive file(s) were not successfully verified                                      BA           195          E
03/11/2014 05:34:43 Job cancelled after system exception ERROR_MESSAGE                                    00           564          A

[Former Member]

thanks for the replies

[Former Member]

Thank You all for taking time and replying .

this was actually an old file which was already moved to storage, hence the program issued error as it could not check the data and proceed with deletion.

Are there best practices to fix archiving errors .

Case 1: the archive program has been cancelld ex: batch job consumed more cpu, time ,so it was killed.

ex: here's the log..

03/10/2014 10:55:03 Job started

03/10/2014 10:55:03 Step 001 started (program S3LIKPWRS, variant 1100_43, user ID ARCH-BATCH)

03/10/2014 10:55:24 Archiving session 052589 is being created

03/10/2014 11:22:43 25 of 109,116 deliveries processed (0 can be archived)

03/10/2014 11:24:14 Path: /archive/x11/zzzarchive/csd/

03/10/2014 11:24:14 Name for new archive file: RV_LIKP_20140310_112414_0.ARCHIVE

03/10/2014 11:52:56 9,550 of 109,116 deliveries processed (816 can bearchived)

03/10/2014 12:22:58 14,075 of 109,116 deliveries processed (1,577 canbe archived)

03/10/2014 12:53:01 17,375 of 109,116 deliveries processed (2,239 canbe archived)

03/10/2014 13:23:04 23,125 of 109,116 deliveries processed (2,664 canbe archived)

03/10/2014 13:53:04 38,375 of 109,116 deliveries processed (2,678 canbe archived)

03/10/2014 14:23:05 49,250 of 109,116 deliveries processed (2,799 canbe archived)

03/10/2014 14:53:07 60,075 of 109,116 deliveries processed (2,916 canbe archived)

03/10/2014 15:23:07 73,500 of 109,116 deliveries processed (2,918 canbe archived)

03/10/2014 15:46:15 Internal session terminated with a runtime error SYSTEM_CANCELED (see ST22)

03/10/2014 15:46:15 Job cancelled

what do we do on such a situation ?

these are daily batch runs for archiving .

is there any note that describes situations and actions for "Archiving sessions with errors"

[JL23]

what do we do?

reading the log

trying to understand what SAP communicates to you.

following the advice given.

To emphasize: the log said you shall execute ST22 to check the run time error (dump) which got created at this very moment, because there is more info and other hints what you can do to check.

[david_liu1]

Hello,

Yes. Jürgen is right.

As per the log you should check in ST22 first.

Regards,

David

[Former Member]

I know how to look into ST22 and all that, thats not the question.

This is not to analyze the log, we know that Basis killed the job, so there's no point in analyzing this.

What is the procedure to fix the failed write session? i.e reprocess archiving.

[JL23]

this part was already answered earlier.

You cannot rescue this failed jobs.

You have to mark your archive as invalid and create a new write job

[gabriela_pieper2]

Hello Juergen,

"You have to mark your archive as invalid and create a new write job" could you please advice me how to do this.

Thanks,

Gabreila

[JL23]

You go into SARA, enter the object, click the Admin button, find your archive run, double click it for the detail, click the pencil get into change mode, activate the box for invalid

[gabriela_pieper2]

Hello Juergen,

many thanks for your quick reply, IS appreciated.

One more question, as I'm not able to check that by myself: are there any dependencies like? Job is started in batch or not, with variant or not, or even the ADK-Job is configured with a test or a productive program.

Many Thanks,

Gabriela

[former_member190120]

Marking an archive session as invalid is not a batch job - it marks it invalid as soon as you save the changes.

Best Regards,

Karin Tillotson

[gabriela_pieper2]

Hello Karin,

thanks for your quick reply. Unfortunately, I'm not quite sure if I really understood your explanation:"Marking an archive session as invalid". Please let me know if this is the way that you are talking:

Best Regards,

Gabriela

[former_member249394]

Hi Gabriela,

Attached is a screen shot of how you can see a session as invalid.

Double Click on highlighted line, edit.

Hope this helps

Damian

[gabriela_pieper2]

Hello Damian,

yes, this help me very much. The session is of course, an "Incomplete Archiving Sessions".

Thanks you again, and thanks a lot all persons involved here.

Best Regard and have a nice weekend!

Gabriela

PS. I didn't used the SCN-Forum for a while, but I'm realy impress for the quite answer for all of you. I would like to add points, but I supose the call must be set to answered from the reporter.

[david_liu1]

Hello,

The problem you encounter is while the verification phase of the deletion process. The ADK is adding a checksum to each data object during the write step. These checksums are verified in the deletion
phase to ensure a correct file and avoid dataloss. In your case the verification is detecting that the checksums at the end of the data objects are wrong.

Please read SAP NOTE 102363.

You could execute the archiving procedure once again with same selection critera. Then a new archive file will be created with same content.

Regards,

David

[Former Member]

Yes i agree with Deepak, but to make sure that the files has been wrongly misplaced during the force close of deletion job i strongly recommend to run the deletion job in test mode, if you get the same error in test mode too then you have to repeat the write,store and delete process all over again.

[former_member188883]

Hi Satya,

It looks like archive file at staging location is corrupted due to abrupt cancellation of jobs.

Suggestion would be to repeat the write+store jobs again and then proceed with deletion later.

Regards,

Deepak Kori