The Future of NetWorker

Thank you for the reply!  Yes, it does autorecycle, and I do let it recycle my incremental tapes.  But, with the full tapes, I need to take one set per month out of rotation because I need to keep that set practically forever.  So, I have about 5 tapes that belong together which I need to find, but which 5 tapes out of the 200 tapes do I need?  That is why I go looking through the tapes looking for the date of the savesets on them.

As far as I know, the recycle configuration does not have a way to handle the odd tape scheme I have to follow.  If you can think of a way to set that up, please throw it at me.

Keeping backups forever is not backup. Nevertheless, if you use backup infrastructure to store your data until further notice time, then you should create separate group and pool for that and run that one as full only on specific dates based on your requirements. This way, you won't mess us with rest of your config.

I'm not sure how your config is done and it depends on that how I would make a search for it.  I would use mminfo to do the query and to give me what I need.  mminfo is powerful command and worth using.

Thanks for that.  Good suggestions.  I hadn't thought about a separate pool.  And, I'll look into mminfo.  I do like a powerful command.

Thanks.

The most awaited enhancement for us is VADP support of file-level recovery of non-windows

vadp clients. Is it being discussed for next major networker release? Is it Networker 8? when will we get a preview?

Thanks!

A feature of Networker called "saveset consolifdation" would be very useful if t was not for its limitations:

1. Save sets cannot be consolidated when either the full or Level 1 save set contains a file larger than 2 GB.
   
2. Renamed and deleted directories cannot be consolidated. If incremental data  contains renamed or deleted directories, the save set consolidation process will detect it and abort.
   

I wonder if this changes in Networker 8. It saves many NW clients from having to be storage nodes!

Julian - Saveset consolidation never satisfied anyone. It will be replaced in NetWorker 8.0 with a new technology - Synthetic Full - that will remove the restrictions imposed by Saveset Consolidation, and will meet the current and future needs of the markketplace.

To quote one of the NetWorker 8.0 beta testers..."Cool new feature!!!!!"

Best regards,

Mark

@Mark,

New DDOS support synthetic approach by NetBackup via Boost so there is data movement by application.  Will the same apply to this new feature in NW8?

Thank you Mark!

I think the single biggest improvement for Networker would be to provide a competent search engine on the Powerlink site.

I get more useful information from the community forums and Google than I do from searching the knowledge base.

I'd echo that one, Powerlink is no longer useful, it only returns a long list of crap.  It did used to work, not particularly well, but you at least had some chance of finding something useful returned.

I haven't been using PL for quite some time as using newer support.emc.com seems to be much better match - especially for search.

Some features I'd like:

• Logging centralized:  Somebody else mentioned having to collect restore logs from each client.  I'd like to have these on the server.
• NMC, logging exposed:  In NMC, once the alerts leave the screen, I can't get them back without digging into log files.  And there are so many log files.
• NMC, group logging history:  Group history would be very nice.  Presently I can see only the most recent run status.
• NMC customization:  In 7.6 you added customization to NMC, allowing font changes, arranging windows, etc.  These settings go away each time I restart NMC.  Can they be save to my account?
• NMC, multi object edit:  I'd like to be able to highlight several tapes volumes and recycle them; highlight several groups and disable them; etc.
• Better alerting:  Alerting is too chatty.  For instance, if a group  is waiting for a device to come available, I get an alert email.  Waiting for a device is normal operations and I shouldn't be alerted to this.  R.Friberg has a good idea that would affect this:  Don't poll the device every 5 seconds.  Instead the jobs should be smarter and determine that the device is just busy.  If there are no jobs using the device, and this job still can't get the device, let me know.
• Stability:  Is anybody else seeing save.exe and savepnpc.exe crashing on Windows clients?  I have tried versions 7.6.1 through 7.6.4 on my clients, and they all crash during save operations randomly.  With retries I eventually get a good backup.
• NMC performance:  Just generally faster would be good.

• NMC, multi object edit:  I'd like to be able to highlight several tapes volumes and recycle them; highlight several groups and disable them; etc.

That's now introduced in NW8, where depending on the column you check, it will be editable, for instance adding the same comment to multiple savegroups and as a next step disabling them.

• Better alerting:  Alerting is too chatty.  For instance, if a group  is waiting for a device to come available, I get an alert email.  Waiting for a device is normal operations and I shouldn't be alerted to this.  R.Friberg has a good idea that would affect this:  Don't poll the device every 5 seconds.  Instead the jobs should be smarter and determine that the device is just busy.  If there are no jobs using the device, and this job still can't get the device, let me know.

Would be nice as more or less we aren't using alerting anymore in the traditional way. depending on size of environment based on "normal amount of fatal errors" we've implemented a counter that only reports an issue if there are more than 10 or 20 or even more media fatal alerts. But that means you might miss out sometimes on a real issue as it gets drowned in the shear endless amount of fatal media errors.

Back in the days up until NW6, we reported any media fatal alert as then they really pointed to an issue. Now fatals are reported much to easy by NW since NW7.

For instance a storage node that has 4 dedicated tape drives, reported multiple fatal media alerts for a SAP backup that runs with 8 sessions with 2 sessions per avaialable tape drive. So even though enough drives where freely available, NW found it necessary to report alerts a couple of minutes after backup start, somewhere between mounting a tape in the 3rd or 4th tape drive. By the time you're logged in to check, the backup is nicely running on 4 drives with 2 sessions each. SO nothing wrong, but still 3 or so pointless fatal alerts.