Expert assistance only please: need serious assistance in understanding BE12.5

[macinslaw]

I have written several posts looking for spotty help here and there.  The real issue is this:

I have never had to use backup exec for more than a simple backup plan until now.  We have a complex routing that we need to implement, VM Ware servers, both enterprise and free, backup to disk to tape and I need to truly understand what is going on better under the hood with the settings that I see.

I know this is a project, and that this is really a training issue, but we have no real budget fr training, and Symantec's Instructionals, KB's and tutorials are either written by someone wanting to show off their vocabulary and grammatical prowess, making them difficult to follow, or are wrong (or just plain don't work).  I am running into the "doesn't work" catagory with Enterprise Vaulty at this very moment.

So I am turning to you to aid in helping me understand how to get the most from BE.

Scenario:

I have a selection list of all my servers and data.  I have created a tiered backup strategy:

daily incremental restet bit,
Weekly - Full reset bit,
Monthly - reset bit,
Quarterly - reset bit
Yearly - reset bit.

Dailies run every day and are kept for one week.  The weeklies run and are kept for 3 weeks until a monthly is run.  When the monthly is run the weeklies become recyclable.  When the quarterlies are run, the past monthlies are now moot.  Yearly, replaces all.

We are running BU2Disk now.  What I am having trouble with is this.  Retention.  we have a 12 Tb drive array.  After the first week we were down to 1.5 Tb.  The data is not being recycled.  I need help to get this set to recycle the various folders on teh array at their designated time.  Also, after the weekly is completed, I want to backup only the data that just completed, ie if a weekly ran on Sat.  Then I want that weekly to be written to tape, if a monthly ran, then the monthly needs to be written to tape.

Right now, I have backup overkill going one.  I have some servers in the restore directory that are just the server name and then I'll have another entry for the servers with the domain tacked onto their name.  I believe these are drive snapshots.  I need to know why I have both, and whter one would suffice.  Also, I need to understand what the .dr is and how I can use it.

I have been able to attach to certain VMware servers, but I cannot run a backup against our vmware cluster.  As a side note, I also have a free vmware server I need to backup the virtual machine files and cannot.  Then there's the fact that I cannot get a good exchange backup because I am getting an error that BE cannot connect to the log files.

WHEW!

See?  A PROJECT!

Can someone assist me?  I have been fighting with this for a month, I've read every article I can get, scoured this forum, but still I am unable to glean much from these methods.  Sometimes you just need someone to help you personally and slap you with the info!

Thanks for your help in advance (that is if anyone is brave enough to try to assist!) :)

[macinslaw]

NO ONE?!  There's no one here who knows how to do these things?

Okay, Thanks anyway.

[Philip_D]

"We are running BU2Disk now.  What I am having trouble with is this.  Retention.  we have a 12 Tb drive array.  After the first week we were down to 1.5 Tb.  The data is not being recycled.  I need help to get this set to recycle the various folders on teh array at their designated time."

Do you have the BU2D media in media sets with the appropriate overwrite protection plan set?

[macinslaw]

Yes.  But the media continues to append, until the drive is full, then starts to overwrite.  I need the backup area to stay clean and lean without my intervention, currently I am having to perform the weeding myself.  Also, if there is any way to get the backup to tape to only perform a backup on the data that was newly written ignoring the elder stuff that would be great.

[Philip_D]

So what you're saying is that your media reaches the end of its overwrite protection period and subsequent jobs are still not able to overwite?

"Also, if there is any way to get the backup to tape to only perform a backup on the data that was newly written ignoring the elder stuff that would be great."

That's exactly what your daily incremental backups are doing.

Incrementals are like differentials of differentials.  They back up only the changed data since the last full backup OR the last incremental backup, whereas differentials back up all of the changed data since the last full backup, even if it's been backed up already since the last full and hasn't changed again.

[macinslaw]

Right...to Disk, not to tape.  Then I run a tape backup of the disk location..........(ding)...... are you saying that I need to set the tape backup not to full, but to differential on the backup disk location?  Would that take care of it?....hmmm....hadn't thought of that.

No, what I'm saying is:

I have a backup that runs, collects data, places that data to a disk array on the network, assigned to a media group.  That group is set to only 1 day, then is supposed to overwrite. However, when the next job runs, it does not overwrite, instead it adds more files to that disk array.  When the disk array becomes full, then it begins the overwrite.

[Paul_Gazo]

If I'm following (and that's quite the wall of text) correctly, what we need to know is what you're backing up and how.

Specifically are we talking about something like Exchange and using GRT?  If so, you've got a lot of reading for that topic.  GRT B2D backups do annoying things.  Painfully annoying things.  Basically they don't obey media recycling rules because, according to Symantec, blah blah these are not the droids your looking for blah blah it's better if it doesn't work the way you expect it to blah blah ignore the man behind the curtain blah blah blah can I have a pony? 

What happens is IMG###### folders are created and never go away until the space needs to be reclaimed.  If a B2D device contains both GRT and non-GRT data, this can eventually cause all kinds of problems such as wishing a meteor would obliterate Symantec's headquarters.  Oh, and filling the drive entirely and making the B2D device offline until you manually sacrifice a chicken and delete the folders.

Maybe your problem is unrelated to this.  Also, maybe I'm over-reacting, but it's been since BE11d now and B2D GRT is still obnoxious.  Yes, it's manageable and yes there are workarounds, but expecting it to behave in a rational manner won't get you anywhere.

[Symanticus]

haha.. yes, that's true mate, at the moment I am struggling with keeping one server per B2D folder.

[Colin_Weaver]

Some comments:that might help:

If your media sets are configured with Infinite Append then they will continue to be used (unless your jobs are set to start with an overwrite) until the media itself is full. Every time a single piece of media is used (in the case of B2D backups) the Overwrite protection settings move back as such if you have an overwrite protection setting of 1 week (intending to overwrite the media the next week, but allow the media to be appended to thoughout the week, it will not overwrite until one week after the media itself is filled or last used.- this will result in media not being overwritten when you expect.

Also bear in mind the overwrite protection is based on the end of a job - so to allow a piece of media to be used at the start of the same job one week later, you actually have to set the protection timings to 1 week less at least however many hours the job takes to run - otherwise again the media will not be overwritten when you expect.

With regards the GRT IMG file comments - enhancements have been introduced in the latest versions of Backup Exec so that every GRT job attempts to delete at least one IMG folder (as long as at least one folder is present that is outside of its overwrite protection period.) This is independent of how full the volume is.

This is a change from the process described by Paul Gazo above which did apply GRT processes in older versions of Backup Exec where IMG folders were olnly reclaimed as the B2D location neared a space threshold.

[macinslaw]

post error.

[macinslaw]

Gazo - you are a tech like me!  I love your response!  Actually, your post describes much of what I'm seeing, so they might be the droids I'm looking for!

albert - Sorry for your troubles, all I can say is this whole software package has gone nuts!

Colin - The whoe issue that I manually set all overwrite protection even as low as 1 hour and all my jobs for this month are still there. 

Yes I am backing up the following:

Exchange - although cannot connect to the information store according to the logs for unknown reasons - and using GRT.

VMWare Enterprise servers, now here's the kicker...the servers that everyone said I wouldn't be able to backup, clustered, are backing up fine, its the single server in vmware that I can't get to backup. - Also GRT.

Regular file servers and domain controllers, 2 sql servers and a partridge...nevermind!

All are backed up over a weekend with a full backup, incremental during the week.  (Yes I know another wall of text....but that's how I roll)  No matter what I do neither teh img folders neither the backup filesare removed, and yes, like Gazo said the backup will cease functioning if the D2D drive is full until I manually remove backups from that drive, so it won't even overwrite....this was confirmed withing the last 2 weeks or so.

HEEEEEELLLLLLPPPP!!! :)