I have just updated my bug list with a new one to report which I have finally solved myself, despite having support request open for 15 days. It was a UI bug :(
So I just logged in to check on tonight's backup. Spoiled again. This time I have got End Marker Unreadable following a completely different job (I had marked the job I had suspected before as Lowest priority to try and fix things up). But there are no job failed errors.
This is getting ridiculous. If my company's main building burned down tonight, they would be well and truly screwed. And I'd be out of a job as a result. Upgrading BE is like playing with fire.
and it was driving me nuts!.....kept on getting tape head errors and saying the heads needed cleaning on my Dell Powervault TL2000. My hardware will usually show an amber light if the heads are dirty. It was not showing.
I then tried a brand new head cleaner.....same errors. After lots of searching I found this KB and it got rid of my errors with the tape drive library. It seems the upgrade messed up something in my SQL DB.
It may not work for you, but thought I'd pass it along.
So I installed this tonight because our backups were flaking at one client. I figured it'd be another incremental upgrade like 11d to 12 or 12 to 2010. Boy, was I wrong. Yeah, due diligence, yada yada yada. We don't have a test environment. I guess this client became one.
And given this experience, they will be the only one. I have no substantive issues with the UI. It will take some getting used to.
However, the server-centric view is terrible for us. We use tape. We have to. Oh, sure, for this one client it's not a big deal. They only have two servers. It didn't take long to update the jobs. But for our clients with more complex setups, upgrading would be an unmitigated disaster. Our clients pretty much only have one tape drive. With this new method, we'd need to schedule up to twelve backups (maybe more) a night for some clients where they have two or three now (or even one). The administrative overhead for managing these jobs becomes a nightmare. Our restore firedrills become a huge ordeal. Yeah, we'd increase the billing for them to account for the longer process. But that would price the firedrill out of reach for some of our clients. And when those clients inevitably have a restore issue, there's a whole lot of finger pointing and blame. No one wins.
At this point, most of our other clients are up for Backup Exec renewals. I don't see us purchasing any. We'll probably ride out 2010 until we can find a software package that meets our needs as Backup Exec 2012 does not.
SDR or Simplified Disaster Recovery - a complete nightmare.
Simplified it certainly has but I think someone at Symantec may have misunderstood the brief.
Simplified was not supposed to be to take away and break a load of functionality.
So far I have spent over a week in time trying to get it to do what IDR did very well.
For the record I have worked with every version good and bad over the years from BE and have been through this type of pain with duff versions before but nothing quite this bad.
With IDR it was quite straightforward to do a bare-metal recovery of all your servers in the event of total loss of access to your LIVE site building and any hardware as it is all destroyed or inaccessible.
Let me work through a scenario to show the types of problems encountered:
So you create the SDR DVD/CD on the 01/03/2012 at 10am
Then you need to invoke Disaster Recovery (for whatever reason) say on the 12/12/2012 at 14:30. You do not have any servers built.
When you try to recover the first server (a lightly important server this being as it is the Media Server) you cannot because the DVD/CD only knows of the media created before the time of the SDR DVD/CD creation.
Symantec say to do this you must have a copy of the .DR file – and they say to get this from the server on the LIVE site !!!
So to make this work they suggest you need to put a copy of the file on say a USB device every time your backup jobs run.
If you run a month turnaround of tapes then following their advice you will need 31 USB devices to keep with the offsite Media and every time you use that Media you must update the corresponding USB device.
Eventually when you manage to run a Disaster Recovery on your Disaster Recovery Kit for the first time and then for some reason you need to run it again the system restores without any errors but when you reboot the server there is no boot MBR on the boot disk and windows repair utilities cannot find any Windows OS install on any devices. For some reason it will not recover the OS to the same box a second time.
To fix this you if you are running HP servers you must run smart start and run the system erase utility - and yes you must delete all temporary logins. Then it might work ok, but you might need to repeat the above until it works.
At the end of all this if you have not lost the will to live and you boss has not killed you for the delay in recovering the servers then you will have recovered your media server and now can recover the rest of the servers.
I have raised this and more with Symantec. We have had long discussions going nowhere on the phone to persons in their call-centres and been sent various videos to watch – which were no use what so ever. Our problems have been escalated all the way up but so far nothing.
We are currently not upgrading any of our customers to 2012 and will in the absence of a workable version (because it is more than a fix that is needed) will start to look at upgrades. It’s a real shame because 2010 IDR worked pretty well for us and our customers.
This worked for me an a few others (I posted this in an earlier thread)
I was experiencing the same message about the deduplication option. In my case, I had once installed it as a trial. Like others, it no longer appeared in my installed options. I made sure the PureDisk service was disabled then removed my backup-to-disk folder and finally removed the registry entry for HKLM\SOFTWARE\Symantec\PureDisk. I restarted my media server and was able to perform the upgrade to 2012.
What is going on here!?! This is the worst product on the market today. What was Symantec thinking... and who Beta tested this?
So, I've ranted about the inability to excludes dats for certain jobs, so I will only minorly touch that subject on this new issue for me. I write this regarding only one of my many new jobs. In this case, backing up my Symantec EnterpriseVault.
For this, I have 4 jobs... one for my daily incrementals, one for the typical end of week, one for the end of month, and the last for the great end of year backup. The first two write to disk, the later two job write directly to tape (If someone chimes in and tells me to send all 4 jobs to disk, then you can purchase my client a new disk subsystem to hold all there backups).
Because I cannot exclude the dates in the calendar, my end of year job run will also include the end of month and end of week backup job, thank you Symantec. Each job takes 14 hour to perform just on this one particular job. 28 wasted hours that I can no longer perform any other server maintenance, or archive emails into EV.
Due to the above issue, I setup all end of year with highest job priority, end of month to high, etc.
Well, end of month kicked off last Friday. It did NOT run based on priority. And, seems my end of month backup wrote to the disk, and not to tape as specified in the job. In the end, it backed up in reverse order which caused my backup disk to run out of space, and no monthly backup jobs completed sucessfully.
I am done. I'd call Symantec Tech Support, but they cannot write me a brand new product. There are just too many issues. I'm not even going to waste my time and energy. In the same time frame, I can roll this back to BE 2010 and see if anything comes out of this Symantec event in two weeks.
I am going to the "Symantec Summit" at Florida. I am really hoping that the Product Development team takes what we have to say very seriously, as this could force us all back to Arcserve (please noooo)!
Every blog that I have read on this site appears to be complaining about the same thing the HORRIBLE user interface in Backup Exec 2012.
I consider myself an expert with Backup Exec and have used it extensively for 8 years. From version 7.5 to 2012 with 2012 being the absolute worst!!!
The interface change is just exactly like others have stated... it is like the Office 2003 to Office 2007 upgrade. Everything changed so much that avid users couldn't work. Now Microsoft corrected many of the issues with Office 2010 by making it more like Office 2003. So hopefully Backup Exec will do the same!
I have always been a fan of Backup Exec products. They have had their issues like most every application but overall have been very stable and worked without issue. Oh and you didn't have to watch a webinar to learn to get around in them.
I have installed BE 2012 on about 6 of my clients and everyone of them have been a pain in the @ss to configure, troubleshoot and of course little to no tech support was available for the product since it is so new. SInce I bill by the hour I have had to discount some of my hard work due to having to spend more time trying to perform the most simple tasks I have performed for years...
In my professional opinion unless you have a lot of time to train on 2012 you better stick with 2010 R3 and wait for the next release which I can only hope will be much better and return some of the features that were in previous versions!
This version... Backup Exec 2012 is just DUMB by design.!!!
I am afraid that the same kind of wisdom and cost cutting measures that produced the wonderful items like the iPhones short little charge cord (which was developed by iDumb and iDumber) was tapped to make major changes to Backup Exec, a product that WAS pretty solid and now is COSTLY GARBAGE!!!!
We are a Symantec Partner and most all of my fellow engineers are very disappointed in the new release!
To be fair the "Convert to Virtual Feature" in Backup Exec 2012 is the bomb and is VERY useful but other than that this release overall just plain sucks to work with!
I think the backup scheme you describe is as follows - correct me please if I don't have it right:
* Daily incrementals to disk
* Weekly fulls to disk
* Monthly fulls to tape
* Yearly fulls to tape
To do this in Backup Exec 2012, you should have 1 backup for the protected server. That 1 backup would have 4 backup job tasks within it - one for each of the above (You can add additional fulls in the Schedule tab of the backup options). If you schedule the monthly full and yearly full to be at the same date and time as the weekly full - then only the least frequent one will actually run. Suppose your monthly full was set to the last friday of every month at 10pm and your weekly full was also set to every friday at 10pm. On the last friday of the month, only the monthly backup will run "superceding" the weekly.
Also, as an alternative configuration, you could have only your daily incremental and weekly full as your backup job tasks. Then add 2 duplicate stages from that to duplicate to tape for your monthly and yearly duplicates. In each of those duplicates you could set it to pick up "the most recent full" as the backup to duplicate to tape. This would avoid having to touch the client machine twice and would still allow you to have your weekly full available on disk for restores.
Concerning the issue of your backup running to disk that was set to run to tape - this should not be happening and I have not heard of it. I'd like to hear more if you are seeing some situation where that is happening.
Mine's a clean install of 2012, not an upgrade. More spoiled backups later and I can confirm there is no pattern to this. Sometimes I get End Marker Unreadable on the tape after RAWS backup, sometimes AVVI, sometimes ADBO. And always the job which precedes this issue finishes cleanly with no errors.
So in summary, it's just very unreliable.
Just discovered another bug if you exclude any file on the C drive.
IE a notepad txt file in a TEST directory call mytest.txt containing the word test doc.
Then SDR green light goes out and it claims the backup is on longer backing up all the required files for SDR to work !!.
Show us where, in the advertising and purchasing processes with third-party vendors, Symantec has informed anyone that this product is a "totally new product with interface changes and the way it works"
This is outlined in the requirements and recommendations for running and creating and SDR. Exclusion of anything on a critical system drive does in fact render the SDR light to turn "off".
For me the same issue remains - this change makes more work. Time is not a luxury I have. I'll wager nor does anyone else here.
Currently I have a handful of servers and ended up with multiple jobs. At a past place of employment I had 52 servers. Under this new model if I had 3 jobs per server (which is what happened when I upgraded) I'd have 156 jobs to manage! Even if I only had one job per server 52 jobs is still not acceptable. This increases the complexity and timing of backups. Complexity breeds issues like rabbits.
As mentioned by others it takes backing up up my data from being a minor activity to a major time expenditure. Neither something I can afford nor burden another resource with, I need them focused on strategic things that bring value to the organization.
Thus I will be forced to find another vendor when my support agreement is up.
"Setting the order that jobs start may be problematic .. the only way I can find is to use the priority setting. There are 5 priority settings to choose, which may be enough for most organisations."
In most cases, as long as your Overwrite job starts first, does it really matter what order the remainder of the job run in?