Unsolved
This post is more than 5 years old
5 Posts
0
1920
Adv File Problem : why NW delete save set for recovery space automatically ?
Hello guys,
First sorry for my english, but I try to explain my case so you can understand :-)
- I'v NetWorker Server 7.4 sp4 run on WinServer2003 r2 32bits, with DiskBackup Tier 2.
- Since we have this configuration, I've something trouble me and didn't know why it does it, and how to fix it :
I backup my servers :
- make a full backup on LTO media in order to externalize theses media in case of disaster. One per month.
- make a incremental backup on adv file media the rest of time, every night, in order to restore more faster. ( we working in pre-press, and have lot of modifications in pictures, texts and sometime (manytimes ;-) ) we have to restore older version of image etc...
And here is my trouble :
- Every morning I checked if backup is successfull, I see if it printed on my LPR the bootstrap info after backup, see in NMC successfull group.
But in Log I see this :
- Media info : 247 Gb recovered by removing 39 save sets from unit F:\SVGDISK.01\_AFreadonly.
- Media info : 23 Gb deleted from save set 15836671 on media F:\SVGDISK.01
- etc....
- F:\SVGDISK.01\_AF_readonly space recovered from media SVGDISK.01.RO
and it do it for all my others media SVDISK.01 to 06
- don't know why it recovering space, because it have space available ( my partition is a 8 To capacity, and there is more than 2 To disk space)
So when I want to restore files from my backup, sometimes I can, because it did'nt delete the saveset with the file, or I can't retrieve the file because NetWorker delete the save set for recover space..... but why it do it !! don't know.
I didn't see where I can fix it in Networker, what is the configuration file that tell NetWorker to recover space ....
So if someone have some ideas about it, I take it !
Thanks in advance,
Chirstian
Polska1422
128 Posts
0
October 7th, 2011 09:00
Hi Christian,
The AFTD deletes expired save sets depending on the retention policy and the recycle setting. Can you check your settings about these policies?
Regards
Claudio
ble1
14.3K Posts
0
October 7th, 2011 13:00
When you save some data, this data is saved and kept certain time. Time used to keep this data is controlled through two settings:
- browse policy - how long you keep data browsable
- retention policy - how long you keep data on backup media (tape or disk)
From above, browse policy <= retention policy.
Now, when retention policy has passed, NW will run cleanup and remove this data from disk.
Both browse and retention policy are set on client (and retention policy can be overrided by retention policy set on pool level).
Example: I have client called clientA and I backup folder PICTURES_XYZ with browse and retention policy of 1 week.
After one week, this data will be removed from disk. (ok, in reality, there is also something called backup cycle which manifestate itself in dependency of incrementals upon previous backup level until full backup, but this is less important now).
chroma3
5 Posts
0
October 18th, 2011 06:00
Hi,
Thx for your answers. I'v changed my browse and retention policies.
First my browse and retention policies are the same : SIECLE (I've create it 100 years)
maybe it was not a good idea to keep it too long, and have the same policy for browse and retention ?
Now : I fix browsable for 10 years, and recoverable for 20 years for just one client, kee others to SIECLE/SIECLE
I'v checked my reports, and it's OK for my client with new policies, my saveset are no longer removed by nsrim x.
I'v checked several days, all works fine. Maybe fix the problem.
ble1
14.3K Posts
0
October 19th, 2011 06:00
I doubt you fixed anything as your change is not applicable to what you see. Just imagine for a moment: you have retention 10, 20 or 100 years and you see those savesets removed? That would mean you already had them there for 10, 20 or 100 years, right? I doubt you use NetWorker for such a long time. Maybe it is not years after all?
chroma3
5 Posts
0
October 19th, 2011 08:00
Hi Hrvoje,
Maybe I doesn't explain as well what I'm doing to fix the problem :
- What is OK now : client with retention policy : 20 years browse retention : 10 years
- What is wrong : client with retention policy : SIECLE browse retention : SIECLE
Clients are in the same group, so when the group finished : it remove saveset from client with SIECLE, and no more from client with retention policy:20years and browse retention:10 years.
SIECLE was a test, and it was years in period (not minutes or days, weeks,months), number of period : 100
Christian
ble1
14.3K Posts
0
October 20th, 2011 04:00
What you say still doesn't make sense, but ok. The original question was why do you see this removed from disk and then you said you can't restore it if it is removed. I doubt it is removed if you use retention periods as you stated (it does not matter if it is century or less up to 10 years). What most likely happens is that ssid is staged and this is why space is recovered. In such case, restore should work fine, but it in return it should request tape where data is staged. Of course, what is possible, is that you have a nsrstage job which stages save sets with retention > 10 years so you picked up those which had incorrectly 20 years before... (I somehow doubt that within 10 years from now you won't fill up that device though).