Of course you should choose the save time according to your needs.
Then you must check manually for all save sets. NW will report failed save sets but it does not know whether some are missing due to your configuration, false directives, removed/lost access paths to disk etc.
I need a list of clients for which backup data is not available for last 10 days or the last backed up data available is 10days older. Is it possible by any chance in networker.
Indirectly - simply use mminfo oncel again as described. Just change the time range to 10days. Then simply verify the output - if a client does not have a save set, it is obviously failing.
Of course you can verify each savegroup report - there you will see a failing client earlier.
The easiest and also the cleanest way is via DPA on the backup report card. But of course will require DPA to be setup and used.
On Networker have you tried the reporting functionality via the NMC?
On Bingo's solutions, it is a very intensive solution. One which our client has used before, where someone has to go through each group/owner's notification email report and verify if a backup succeeded or failed. The mminfo method would require someone with an understanding of what is the normal success state and failed state. IE. System A is meant to have saveset C:\, D:\ and SYSTEM STATEs. If missing one then its a failure. Doable but very intensive as well.
Between the two intensive methods I would go with the savegroup/owner's notification report, the reason is that the individual report will allow the system administrators to know the state of each saveset and also whether it is a real success or a "fake one". IE. Have seen in the past where the SYSTEM STATE has a warning issue with backing up Cluster information but still reports success for the saveset. You wont catch these sort of errors with the mminfo method.
DPA is the easiest and cleanest way but it costs, unless there is a NMC reporting function that meets your need (one which I havent really used at all as we have DPA).
In general, i think a good idea would be to detect abnormal changes in size.
Query/report 'client & save set & level & sumsize' and compare the size with the last instance of the same backup. If the difference is more than x%, trigger an appropriate notification. There is at least one enterprise BU solution which offers this feature.
I need a list of clients for which backup data is not available for last 10 days or the last backed up data available is 10days older. Is it possible by any chance in networker.
This can be easily scripts. There are few steps:
a) obtain list of clients via nsradmin
b) use list from a) to run mminfo query for past 10 days
c) if last backup data is >= 10 days report name of the client
If retention used is more that 10 days then you could use mminfo to obtain list of clients. Caveat is that if you added new client, let's say, 5 days ago and never had successful backup due to probe failing you would not see that client in list reported by mminfo as such backup entry would not exist. Similar, if you rely on solely nsradin outup, if someone removed client 5 days ago by mistake, you would not see that client listed in step a). For paranoids, you can combine both methods and sort by unique name to avoid such rather rare and unlikely situations.
bingo.1
2.4K Posts
1
June 22nd, 2011 22:00
The only way to do this is by querying the media index ... for EACH client.
The syntax would be something like
mminfo -q "client=name,savetime>'1 month ago' -r "client,name,savetime,level" -ot
Of course you should choose the save time according to your needs.
Then you must check manually for all save sets. NW will report failed save sets but it does not know whether some are missing due to your configuration, false directives, removed/lost access paths to disk etc.
bhargav_tarigop
17 Posts
0
June 23rd, 2011 17:00
Let me ask the question in the other way.
I need a list of clients for which backup data is not available for last 10 days or the last backed up data available is 10days older. Is it possible by any chance in networker.
bingo.1
2.4K Posts
0
June 23rd, 2011 21:00
Indirectly - simply use mminfo oncel again as described. Just change the time range to 10days. Then simply verify the output - if a client does not have a save set, it is obviously failing.
Of course you can verify each savegroup report - there you will see a failing client earlier.
d0c294551025416
6 Posts
0
June 23rd, 2011 23:00
The easiest and also the cleanest way is via DPA on the backup report card. But of course will require DPA to be setup and used.
On Networker have you tried the reporting functionality via the NMC?
On Bingo's solutions, it is a very intensive solution. One which our client has used before, where someone has to go through each group/owner's notification email report and verify if a backup succeeded or failed. The mminfo method would require someone with an understanding of what is the normal success state and failed state. IE. System A is meant to have saveset C:\, D:\ and SYSTEM STATEs. If missing one then its a failure. Doable but very intensive as well.
Between the two intensive methods I would go with the savegroup/owner's notification report, the reason is that the individual report will allow the system administrators to know the state of each saveset and also whether it is a real success or a "fake one". IE. Have seen in the past where the SYSTEM STATE has a warning issue with backing up Cluster information but still reports success for the saveset. You wont catch these sort of errors with the mminfo method.
DPA is the easiest and cleanest way but it costs, unless there is a NMC reporting function that meets your need (one which I havent really used at all as we have DPA).
GL.
bingo.1
2.4K Posts
1
June 24th, 2011 00:00
In general, i think a good idea would be to detect abnormal changes in size.
Query/report 'client & save set & level & sumsize' and compare the size with the last instance of the same backup. If the difference is more than x%, trigger an appropriate notification. There is at least one enterprise BU solution which offers this feature.
ble1
4 Operator
•
14.4K Posts
2
June 25th, 2011 07:00
This can be easily scripts. There are few steps:
a) obtain list of clients via nsradmin
b) use list from a) to run mminfo query for past 10 days
c) if last backup data is >= 10 days report name of the client
If retention used is more that 10 days then you could use mminfo to obtain list of clients. Caveat is that if you added new client, let's say, 5 days ago and never had successful backup due to probe failing you would not see that client in list reported by mminfo as such backup entry would not exist. Similar, if you rely on solely nsradin outup, if someone removed client 5 days ago by mistake, you would not see that client listed in step a). For paranoids, you can combine both methods and sort by unique name to avoid such rather rare and unlikely situations.