Networker and tapes expiration

This is because a tape expires by the date as per the retention period of the save sets backed up.

By default, backup server has a retention policy of 1 year and indexes are backed up on the same tape on which backup is taken. Index is a backup server save set. So at end of the backup, indexes are backed up on the tape with one year retention as set for backup server. Therefore, the tape expires after one year even if other client's save sets are having two weeks retention period.

Ways to resolve:

1. Reduce the retention period of the client resource of backup server to be equal to or less than the retention period of the clients in the same group.

2. Configure your backup setup such that the indexes are backed up on a separate dedicated media other than the media on which client save sets are backed up. This way you can have index retention of one year although if the backup media (on which client save sets were backed up) lasts for two weeks only then there is no point retaining indexes.

2. Is it possible to save index each time on the same media where the backup is done? If i lost tape with Index I would have a problem with DR and restoring some data

You can check this by checking the save sets backed up in the volume. You will find a saveset "Index:" with backup server as a client there.

How can I do this?

I can create special pool and tapes. But where can I configure it to save indexes on separate pools?

Ok - I will do so and see if this helps during next night backup.

One more thing. Can you be so kind and shortly explain me what are and why do we need Index and Bootstrap?

What happen if I lost both of them? How long shall I keep these backups? One year is enough? My longest backup should be keept for one year.

Thanks a lot for your help!

This article should help you with that:

How to save the index to a separate pool

http://solutions.emc.com/EMCSolutionView.asp?id=esg54449&usertype=C

-Bobby

Create a new pool with savesets INDEX: & Bootstrap. Mark all the required groups in this pool and label media for this pool. Next time when the group backup is over, it will pick up media from this pool for index backup.

Index is the catalog information of the files that have been backed up. Thsi is stored in your /nsr/index folder and the clientname sub-folder. this helps you track what all has been backed up and when you want a browsable recovery, Networker refers to this index to give you the file-by-file view like Windows Explorer. This is important to backup so that if you have lost your backup server and want to rebuild it, this helps you rebuild the indexes. Also, when you take incr or diff backup, Networker refers to this index information to get details on the last backup attributes for each file.

Bootstrap holds the configuration information i.e. the /nsr/res and /nsr/mm folders. Again, if your backup server crashes, you can install fresh Networker software and recover bootstrap i.e. res and mm folders and you do not need to reconfigure anything.

In event of a complete crash of backup server, recovering bootstrap gives you /nsr/res and /nsr/mm and recovering index gives you /nsr/index folders without the need to manually reconfiguring your environment.

By default, these are backed up at the end of every group backup. Bootstrap is done once a day in the group that has backup server as a client. The media chosen for these backups depends on the configuration as discussed in previous posts of this thread. You can also take an extra backup by creating a group with all clients and then manually running a savegrp -O{groupname} command. The pool you specify for this group will be used for these backups.

NO INDEX AND BOOTSTRAP BACKUPS MEANS NO WAY TO COME BACK FROM A BACKUP SERVER DR.

You can manually also keep a copy of these three critical folders but then firstly it is manual (can forget sometimes) and secondly you need to stop Networker services for it.

You can choose retention period based on how long you want them but remember that if retention period of a given save set of a client has crossed, it's indexes will get deleted from the /nsr/index folder automatically since if you do not retain data, there is no point retaining the index for it.

{In the responses, you have the choice to mark as Correct/Helpful for each response you get. This gives points to the participants responding and encourages them to participate more. Mark the responses based on what you think about them}

Hello Again,

Sorry, but your solution does not work

I created tape Pool index and defined all things like in the article, but backup index are still saved on the backup pools - not on index pool and tape

Shall I create some additional client and group to backup indexex on separate pool tape?

My tapes expiration is still one year instead of for example two weeks or one month - depends of group/client configuration....

Thank you All for your help!

I did so.

I am sure, that retention period for the client resource of the backup server is set to 2 weeks or 1 month (different clients and groups).

My backup is running since one week with new settings regarding new pool for index and the tapes are still marked with expiration of one year.

Furthermore - on the Index Pool and Tape there is no data saved since I create it

One more thing... On the Pool properties, Configuration tab there is special option: Retention Policy where I can choose some dates - for example 14 days, day, month etc - if I set it here as I configure client for exaple for 2 weeks - tapes will be set as expired after 2 weeks?

Or will it cause some conflict between Clinet configuration and Tape Pool configuration?

If a tape has been marked for one year expiration, it would continue to be so untill manually changed or relabeled.

Ensure that the retention period for the client resource of the backup server is set low and you backup on a fresh tape, the expiration date would be new. Remember the expiration date of the tape is as per the longest retention period set for all the client resources created for the group and the client resource created for the backup server. Note down the retention periods and start a backup on a new tape and then check the expiration date.

Ok.

I will recheck my backup policy for two different clients and groups during weekend and come back with this issue on Monday.

Thank you very much for your help!

Have a nice weekend

We generally don't set policies on pools because it is easy to manage on clients. It is hard to believe that all retention policies are 2 weeks still the tape expiration is one year. Is it a freshly labelled tape?

Same is the case with the INDEX pool issue, please recheck: the same article works and I see both the above working here. Recheck and let me know then I will re-think about these.