topic Re: Archive Backup System error : ABS_ARCHIVE_ACCESS_FAILED, Failed to access Archive in Operating System - OpenVMS

Archive Backup System error : ABS_ARCHIVE_ACCESS_FAILED, Failed to access Archive

oxygene65_1 — Tue, 13 Apr 2010 09:21:01 GMT

Hi

Does anyone else have to use ABS (Archive Backup System) on Open VMS ?

I've never used this product before, and have had it dropped in my lap to support.

Our backups are failing with the following error...

20:01:17 COORD: Failed to access Archive
20:01:17 COORD: ABS_ARCHIVE_ACCESS_FAILED, Failed to access Archive
20:01:17 COORD: -Line = 1012, File = RESD$:[SRC]COORDINATOR.C;1

I've trawled the net, and the ABS documentation and cannot find a reference to that error anywhere.

If anyone else uses this old fossil of a product and can point me in the right direction this would be gratefully appreciated.

Thanks

James

Re: Archive Backup System error : ABS_ARCHIVE_ACCESS_FAILED, Failed to access Archive

P Muralidhar Kini — Tue, 13 Apr 2010 09:47:37 GMT

Hi,

>> 20:01:17 COORD: Failed to access Archive
>> 20:01:17 COORD: ABS_ARCHIVE_ACCESS_FAILED, Failed to access Archive
>> 20:01:17 COORD: -Line = 1012, File = RESD$:[SRC]COORDINATOR.C;1

This is a generic error message displayed by ABS. To know the exact
cause for the problem, we need the specific error encountered by ABS.
The specific error message will be present in the Save logfile after
the above set of lines.

Please provide the complete save logfile for further analysis.

Regards,
Murali

Re: Archive Backup System error : ABS_ARCHIVE_ACCESS_FAILED, Failed to access Archive

oxygene65_1 — Tue, 13 Apr 2010 09:52:51 GMT

Hi Murali, as requested, I've attached the full logfile that contains the error.

Its quite frustrating as I've spent most of the morning trying to find that error within the ABS documentation that I have available to me. But cannot find a thing.

Thanks

James

Re: Archive Backup System error : ABS_ARCHIVE_ACCESS_FAILED, Failed to access Archive

P Muralidhar Kini — Tue, 13 Apr 2010 10:08:08 GMT

Hi James,

From the save logfile provided, we can see that the specific error
encountered by ABS is "ABS_SKIPMARKS_FAILED".

I found the following information in the ABS operations guide -

Fast Skip Errors
If you receive an ABS_SKIPMARKS_FAILED error there is a logical name which
may be defined at system level which turns off the ABS fast skip methods.
To disable fast skip do the following command on the affected system:
$ DEFINE/SYSTEM ABS_NO_FAST_SKIP TRUE

You can define the above logical and then run the save again and see
if the save runs successfully or not.

I hope you have access to the ABS Operations Guide document ?

Longtime since i worked on ABS, will wait for some folks from ABS to
give more technical information on the above error.

Regards,
Murali

Re: Archive Backup System error : ABS_ARCHIVE_ACCESS_FAILED, Failed to access Archive

P Muralidhar Kini — Tue, 13 Apr 2010 10:19:00 GMT

Hi James,

This is the information that i could find for FASTSKIP -

When tape mount verification was introduced a long time ago, tape
drivers needed to keep the position on a tape. Thus, when mount
verification is performed, the tape driver can position the tape to
exactly where it was before. To keep track of the tape position, the
only method then was to count blocks. So even a skip forward by tape
marks was done by skipping forward based on blocks. When newer tapes
(mostly SCSI based) came out which could report the exact tape position
at any time, this method was still in use. The FASTSKIP was introduced
in OpenVMS 7.1 (I think). It allows the driver to check for the
capability of the SCSI tape drive and, if it can report its tape position,
skip file marks is used instead of skip blocks. This speeds up tape
positioning (e.g. BACKUP append save set) tremendously because most of
these drives have a tape mark directory. So instead of reading every block
and winding down every serpentine on the tape, these drives switch
track/serpentine quickly and move to the tape mark directly.
For example it can take up to 10 minutes to reach the end of a 10 GB tape
with the old method while it takes less than a minute to use the fast
skip method.

Recommendations from ABS -

On drives that do not support the fast tape positioning, you may see
errors such as, ABS_SKIPMARKS_FAILED, Skip tape marks failure

In these cases, define the logical ABS_NO_FAST_SKIP on the node where
the failures occur.

Looks like its the drive that is not supporting the fast tape positioning feature as a result of which ABS is giving the "ABS_SKIPMARKS_FAILED" error.
Defining the "ABS_NO_FAST_SKIP " logical and re executing the save request should solve the problem.

Regards,
Murali

Re: Archive Backup System error : ABS_ARCHIVE_ACCESS_FAILED, Failed to access Archive

oxygene65_1 — Tue, 13 Apr 2010 10:52:11 GMT

Thanks Murali, I will give that a go and let you know how I get on.

Re: Archive Backup System error : ABS_ARCHIVE_ACCESS_FAILED, Failed to access Archive

Deepa R Shenoy — Tue, 13 Apr 2010 11:17:08 GMT

Hi James,

From the logfile provided, I noticed that you are using RDF to access the jukebox/drives.

>> 20:00:21 COORD: Allocated drive RDEV03$MKA100: in jukebox GKA0

Generally when a SAVE/RESTORE request fails, the last few line reflect the actual problem faced. For example, in your case, "ABS_SKIPMARKS_FAILED" is the error.

>>
20:01:17 COORD: ABS_SKIPMARKS_FAILED, Skip tape marks failure
20:01:17 COORD: -Line = 2045, File = RESD$:[SRC]ABS_MEDIA_MGMT.C;1
>>

ABS usually positions the tape to write (in case of SAVE) or read (in case of RESTORE).
>> 20:00:29 COORD: Skipping RDEV03$MKA100: to End of Tape...

At this time, ABS encountered an error. Due to this, the problem ABS_SKIPMARKS_FAILED is reported.

Please try out what Murali has suggested.

If the problem persists, you can check the following logs.

1. RDF Server/Client logs on
- Node where device is connected
- Node where SAVE is running

Client Logfile - TTI_RDF:RDCLIENT_.LOG
Server Logfile - TTI_RDF:RDSERVER_.LOG

2. DBServer Logfile
- MDMS$ROOT:[LOGFILE]MDMS$LOGFILE_DBSERVER.LOG

3. Node Specific Logfile in case the node where you are running the SAVE is a client.
- MDMS$ROOT:[LOGFILE]MDMS$LOGFILE_.LOG

In the logs, check what was happening from time 10-APR-2010 20:00:21 to 10-APR-2010 20:01:21.19.

Also, let us know what type of device, drive are you using. Is the error count on the device increasing with every IO?
Post the above log excerpts for us to look at it.

Regards
Deepa

Re: Archive Backup System error : ABS_ARCHIVE_ACCESS_FAILED, Failed to access Archive

Deepa R Shenoy — Tue, 13 Apr 2010 11:36:33 GMT

Hi James,

I had another look at the log and found that the name of the node is "RDEV03". I mistook that RDEV03$MKA100 is a RDF device.

>> 20:00:21 COORD: Allocated drive RDEV03$MKA100: in jukebox GKA0

This device is connected and is accessible locally to the system.

You do not have to post the RDF log files. Please post the other ABS/MDMS related log files mentioned in my previous reply.

Regards
Deepa

Re: Archive Backup System error : ABS_ARCHIVE_ACCESS_FAILED, Failed to access Archive

P Muralidhar Kini — Thu, 04 Jan 2024 13:40:16 GMT

Hi James,

Just checking to know if the problem is resolved.

Also, refer the following link which says how you can say thanks
~~http://forums11.itrc.hp.com/service/forums/helptips.do?#28~~

Regards,
Murali

[Moderator note: The above link is no longer valid.]

Re: Archive Backup System error : ABS_ARCHIVE_ACCESS_FAILED, Failed to access Archive

oxygene65_1 — Mon, 10 May 2010 07:19:00 GMT

We finally resolved this problem last week. There were a number of causes, but the main culprit was a SCSI terminator ! This was swapped out with a new one and pretty much all the problems we were experiencing disappeared.

However, the advice on tape skipping and defining the "ABS_NO_FAST_SKIP" = "TRUE" logical also resolved one of the problems.

Thank you for all your assistance with this.