2010-05-06 10:48 PM
I'm Paolo and I'm trying to backup my NAS shares with IBM Tivoli Storage Manager using snapdiff API but when a file have the name with special character the backup fails.
The people of IBM support show me this document:
and at the end of the document it seems that to works fine I've to wait a new version of Snapdiff API.
My Data Ontap is at version 7.3.3.
The question is:
Anyone knows if that problem is solved by a new version on Data Ontap or new version of Snapdiff API ???
I have a lot of files with file name with special character and change the name of that files is a problem.....
Thank you in advance.
2010-06-08 04:33 AM
we see same problem (and same KB article from IBM) at one of our customers. They are on DOT 7.3.2. Volume Language is de.UTF-8.
Since no info about this issue can be found in NOW, we are also interested in a solution, if there is any already.
2010-06-08 05:15 AM
we have DataOntap 7.3.3 and at the moment to bypass the problem we use an hidden TSM parameter that is not already supported by IBM that seems works. The parameter is
2010-06-09 12:40 AM
we are running ONTAP 7.3.2 with this problem. I just did a test with this testflag, and got no skipped files. I don't really know if files with special characters have been backed up, or if tsm client just supressed the error messages...
I remember that I saw this testflag some days ago, but don't know where... Anyway, I'm not happy using an unsupported option of which I don't even know what exactly it does. Do you have any idea about this, or maybe you could share your source?
2010-06-09 01:18 AM
to know if you have backed up that files the best way is try to restore it using TSM client, we have test testflag in this way.
Some people of IBM speak me about this unsupported flag but I don't know when it becomes fully supported.
Anyway it's a strange situation.......IBM certify snapdiff that is an API written by Netapp Community(???!!!) but not fully supported by Netapp !!
2010-06-10 11:51 PM
I've restored an old TSM log for information about files, which have not been backed up bevore I added the testflag to dsm.opt. Idea is, to take one file after the other and compare it's status with last backup and reality on the filer.
Unfortunately, we had a change in access permissions on all files for the new Sharepoint, which results in a new full backup.
An other issue is the performance. The new backup is running the THIRD day and has done only 360G! This is unaccepable poor performance. What about your performance, are you happy with it?
2010-06-11 12:50 AM
We have some volumes of 1TB size with more than 1 million files and after the first full backup (maybe 1 day or 2 days) the next backups (using snappdiff) are very fast.
TSM maintain one snapshot of the volume as "point of start" of the next snapdiff backup and save the files modified from that snapshot.
I think snapdiff is very useful and I'm waiting for the certification of testflag !!
2010-06-14 02:31 AM
Ok, I can confirm now that once skipped files are backed up now (with TESTFLAG SNAPDIFFONTAPFAP). I talked to TSM support which told me to add TESTFLAG SNAPDIFFNAMEFILTEROFF to the options file.
There are some more questions about this, but backup works for now.
Good luck to anyone
2010-12-28 07:13 AM
I've had my eyes on the SnapDiff functionallity since it appared in 6.1 as an alternative to the proxy node traversal method. We are located in Sweden, with the extended ASCII
characters used quite exentsivly, the lack of UTF-8 support made snapdiff unusable. At the time (late 2009), we where running TSM 6.1 and ONTAP 7.3.2.
Since then we have upgraded to ONTAP 7.3.3 and TSM 6.2. I've seen information about ONTAP 7.3.3 support in 6.2.2, and that the snapdiff API call had been reworek in 7.3.3, I thought I's have a go at it.
TSM 6.2.2 was released in early december, so I downloded it, stuck on the test proxy node that does the backups and configured everything for snapdiff backups. Ran the baseline backup, so far so good. Popped a couple of new files on the filer and had another run, all good. Created a couple of new files with extended ASCII chars in the name, and voilá, still working !
Below is the log from the TSM client running the snapdiff. TSM Server and Client running 18.104.22.168, Filer running ONTAP 7.3.3.
Node Name: XXXXXX
Session established with server XXXXXXXX: Windows
Server Version 6, Release 2, Level 2.0
Server date/time: 2010-12-28 14:05:14 Last access: 2010-12-28 14:04:09
Accessing as node: XXXXXXXXX
Incremental by snapshot difference of volume '\\xxxxxxxx\system$'
Incremental by snapshot difference of volume '\\xxxxxxxx\users$'
Directory--> 0 \\xxxxxxxx\users$\x\xxxx [Sent]
Normal File--> 0 \\xxxxxxxx\users$\x\xxxx\ö.txt [Sent]
Successful incremental backup of '\\xxxxxxxx\users$'
Directory--> 0 \\xxxxxxxx\system$\test2 [Sent]
Normal File--> 0 \\xxxxxxxx\system$\test2\ä.txt [Sent]
Successful incremental backup of '\\xxxxxxxx\system$'
Total number of objects inspected: 0
Total number of objects backed up: 4
Total number of objects updated: 0
Total number of objects rebound: 0
Total number of objects deleted: 0
Total number of objects expired: 0
Total number of objects failed: 0
Total number of subfile objects: 0
Total number of bytes inspected: 820 B
Total number of bytes transferred: 820 B
Data transfer time: 0,00 sec
Network data transfer rate: 0,00 KB/sec
Aggregate data transfer rate: 0,08 KB/sec
Objects compressed by: 0%
Total data reduction ratio: 0,00%
Subfile objects reduced by: 0%
Elapsed processing time: 00:00:09
// Andreas Jansson - Sweden
2012-02-29 02:25 AM
could you please tell me which steps you made to register the node in tsm and what youre dsm.opt and dsm.sys looks like?
Thx in advance