Unsolved
This post is more than 5 years old
1 Rookie
•
124 Posts
0
2369
Clones continue on save set error
what does the setting in clones of continue on save set error setting actually do?
will it skip the save set and move onto the next save set if there is an invalid save set or volume identifier?
or will it clone the save set irrespective of the save set or volume identifier being invalid?
crazyrov
4 Operator
4 Operator
•
1.3K Posts
0
August 21st, 2014 12:00
What this would do is, whenever an invalid saveset (failed/aborted) or volume it would skip this saveset or volume and move to the next saveset/volume in the list. if this is left unchecked the clone job would fail whenever an invalid saveset/volume is encountered.
boerbokrib
1 Rookie
1 Rookie
•
124 Posts
0
August 21st, 2014 22:00
Hi Crazyrov
if i got this error with this switch enabled.
info scheduled clone alert: Scheduled cloning 'SAP spp CLONE TO 4200 alsoda12' completed. Total 9900 save sets, 9879 Failed.
89017 8/21/2014 1:52:39 PM 1 5 0 2040 1976 0 alsdnw.ialch.co.za nsrd NSR notice Scheduled resource list: NSR task clone.SAP spp CLONE TO 4200 alsoda12 exited with return code 1.
Then can i take it that 9879 savesets are invalid and i will not be able to use them at all?
How would i go about cleaning them up?
I have run nsrim -X manually
bingo.1
2.4K Posts
0
August 21st, 2014 22:00
Before you do anything else, check carefully for the reason - for your own safety.
There must be a reason why 99% of the save set cannot be cloned:
- Your media db query might be wrong
- The source media is not available
- The save sets may have failed or aborted or are suspect
Just to name some potential issues.
You do not want to tell us that you have created about 10k 'defective' save sets and you do not know why. Correct?
boerbokrib
1 Rookie
1 Rookie
•
124 Posts
0
August 21st, 2014 23:00
Hi Bingo
That is why I am checking.We are cloning to a new 4200. I started cloning and we had a lot of save sets go through then it failed. this is about 18 months of data we are cloning.
I then restarted the clone and it reported ever decreasing amounts
38758 8/21/2014 1:52:39 PM 0 0 0 2040 1976 0 alsdnw.ialch.co.za nsrd NSR info scheduled clone failure alert: Scheduled cloning 'SAP spp CLONE TO 4200 alsoda12' completed with errors.
38758 8/21/2014 1:52:39 PM 0 0 0 2040 1976 0 alsdnw.ialch.co.za nsrd NSR info scheduled clone alert: Scheduled cloning 'SAP spp CLONE TO 4200 alsoda12' completed. Total 9900 save sets, 9879 Failed.
How ever we are at a point now where I am getting 9879 completed and 9879 failed. So I am trying to figure this out before I proceed.
38758 8/21/2014 2:53:29 PM 0 0 0 2040 1976 0 alsdnw.ialch.co.za nsrd NSR info scheduled clone failure alert: Scheduled cloning 'SAP spp CLONE TO 4200 alsoda12' completed with errors.
38758 8/21/2014 2:53:29 PM 0 0 0 2040 1976 0 alsdnw.ialch.co.za nsrd NSR info scheduled clone alert: Scheduled cloning 'SAP spp CLONE TO 4200 alsoda12' completed. Total 9879 save sets, 9879 Failed.
Below is the volumes on the pool we are cloning.
SAPClone.001: 2141 GB used, 1351 save sets, appendable, 1351 browsable save sets
SAPClone.000: 67 TB used, 46077 save sets, appendable, 45826 browsable save sets, 251 recyclable save sets
When I run a nsrim -X of the volume i get the below
alsoda11-vip.ialch.co.za:RMAN:SPQ_aemzqhui.183_1, 0 browsable cycle(s), ad hocs
1 browsable files of 1 total, 99 MB recoverable of 99 MB total
crazyrov
4 Operator
4 Operator
•
1.3K Posts
0
August 22nd, 2014 01:00
Run an mminfo against the savesets that you are trying to clone? This way we can ensure if the clone job is behaving as expected.
boerbokrib
1 Rookie
1 Rookie
•
124 Posts
0
August 22nd, 2014 01:00
Hi Crasyrov
what will mminfo show us that i need to lookout for?
boerbokrib
1 Rookie
1 Rookie
•
124 Posts
0
August 22nd, 2014 02:00
Hi Crazyrov
Here is the output attached.
2 Attachments
SAPCLONE001.rar
SAPCLONE000.rar
crazyrov
4 Operator
4 Operator
•
1.3K Posts
0
August 22nd, 2014 02:00
So, you are trying to clone the tapes off the SAPClone.001 and SAPClone.000 volumes to a different volume? like a tape ?
All the savesets on these volumes are valid. All are in state "cb"
crazyrov
4 Operator
4 Operator
•
1.3K Posts
0
August 22nd, 2014 03:00
Correct, rule of thumb is if a volume already contains a copy of one saveset (ssid) then it will not be able to take in another saveset of the same ssid. You could pull out the ssid's on both the volumes and see if all the savesets are there. If not try to manually close a few savesets and see if you can get a better workable error.
boerbokrib
1 Rookie
1 Rookie
•
124 Posts
0
August 22nd, 2014 03:00
HI crasyrov
Yes we are cloning from one pool on the Datadomain 860 sapclone to the datadomain 4200 pool named sapclonepri.
so as long as it is cb it is fine? they should all go across? Unless of course the saveset id is duplicated in which case it will fail.?
bingo.1
2.4K Posts
0
August 22nd, 2014 04:00
Before i start an automatic clone, i would go for a manual one. This allows better traceability of the process and eliminates some sub-processes which might also cause problems.
Run something like
mminfo -q "pool=sapclone,copies=1" -r "ssid" > ssids.txt
Count the number of save sets.
Then initiate the clone running
nsrclone -b sapclonepri -S -f ssids.txt > outfile.txt
Then check the output file for results.
Duplicate SSIDs cannot be created by the backup server. Multiple instances - of course you might have more than one - are distinguished by the combination of SSID and CLONEID.
And it is not "a rule of thumb" - it is mandatory: a save set (even a partially one) can only have one reference to one media.
Luckily, aborted, suspect and recyclable save sets will be removed from a backup disk ...
- on an AFTD right away
- on a DD sometime later during the day (unless manually enforced)
On an AFTD and DD media, the SS will only exist completely on the volume.
If the save set still exists, NW will simply use another writeable piece of media from that pool (if available) or ask for another writable media from the same pool.
boerbokrib
1 Rookie
1 Rookie
•
124 Posts
0
August 22nd, 2014 05:00
Hi Bingo
That does look like a better way but can i ask why you used
mminfo -q "pool=sapclone,copies=1" -r "ssid" > ssids.txt
instead of mminfo -q "pool=sapclone,validcopies=1" -r "ssid" > ssids.txt
bingo.1
2.4K Posts
0
August 22nd, 2014 05:00
You are correct.
As an 'oldtimer', i am simple using to the old-fashioned attributes more often ;-)
boerbokrib
1 Rookie
1 Rookie
•
124 Posts
0
August 22nd, 2014 06:00
LOL Bingo
Not sure you that old. Just I am new to Networker.