Quantcast
Channel: All Data Protector Practitioners Forum posts
Viewing all 10494 articles
Browse latest View live
↧

Re: Data protector Reporting

November 21, 2013, 2:01 pm
$
0
0

Does "Very frequently" mean every report.  If not, how often?

 

Without knowing what your timeframe is, or any other parameters you may be using, lets see what happens when you run the report from the comand line

 

First

 

           omnirpt -report list_sessions -timeframe 24 24 -tab -log C:\temp\report.txt

 

YOu can send the 'log' file to any location that you want,with any filename you want.  Then

 

            omnirpt -report list_sessions -timeframe 24 24 -html -log C:\temp\report.htm

 

I am not so much interested in the content of the report, rather, if the report gives you the strange-looking output

 

Then, lets send the two reports to e-mail

 

       omnirpt -report list_sessions -timeframe 24 24 -tab -smtp [your e-mail address]     Do not include the [  ] characters

       omnirpt -report list_sessions -timeframe 24 24 -html -smtp [your e-mail address]

 

↧
Search

Re: Tapes on disk marked bad - Is there a way to fix?

November 21, 2013, 2:26 pm
$
0
0

Although Bob's comments are refer to physical tapes (and are correct) they also apply to 'virtual' tapes.  Obviously, the chances of a virtual tape being faulty/damaged is lower than that of a physical medium due to the lack of 'mechanics' and the poor status is almost certainly due to an IO failure during a backup.  To check this, identify the sessionID of a backup on some of these poor tapes and check that session for IO failures.  If this seems to be the cause of your virtual tapes becoming poor then it's a fairly safe bet that you can use reset_poor_medium.

 

However, because virtual tapes are deleted once their contents have expired and are not reused (like physical tapes are) there seems little to worry about.

 

Hope this helps to explain things.

 

John

↧
↧

Re: Backup client on DMZ

November 21, 2013, 10:59 pm
$
0
0

Done all of that.

I´ll get no "denied" in firewall log.

Feels like some DNS sort of error, but when I add CM in client hosfile things getting worse..

DP version 8 and Forefront TMG..

↧

Re: Best setting for concurrent backups for backup to disk

November 22, 2013, 12:21 am
$
0
0

if you backup to disk (file library or storeonce software store) you also need to set the media pool to "non-appendable".

 

for mulitiplexing to a physical tape. here you can say about concurrency "use as many as needed but as less as possible". keep in mind, more concurrent stream for backup can increase the backup speed, but it does slow down the restore speed. normaly i start with 4-6 streams. if this is enough to keep the tape in streaming mode (Data Rate Matching for the Device) then i leave it like this...

↧

Re: HP DP and EMC storage - restore error

November 22, 2013, 3:35 am
$
0
0

Hello,

 

no, this is just mount point. Server is booting from local hard drive. I don't use native SLES multipathing, I use EMC powerpath and named device /dev/emcpowera. Via system I have see the same device as /dev/sd[c-o], 12 paths.

↧
↧

Re: HP DP and EMC storage - restore error

November 22, 2013, 3:44 am
↧

Re: Backup client on DMZ

November 22, 2013, 3:58 am
$
0
0

hi tom

 

there are severall things to considure if you want to backup a client in dmz. do you need filesystem backup or also integration backup.

 

  1.  Make sure nameresolution does work on both directions!!! from the cell server, you need to resolve fqdn (dmzserver.domain.local), from dmz server, you need to resolve fqdn from cell server (cellserver.domain.local). if you have a dedicated media agent, this host needs do resolve dmz server and cell server as well. i think cell server is also media-agent?
  2. you need to open tcp port 5555 (default dp communication port) from cell server to dmz server.
  3. you need to open tcp portrange for "backup stream data". first on the media agent, first copy omnirc.tmpl and then rename the copy to omnirc. (if you not already have this) then go to the line "ob2portrangespec". modify with a port range. for example:

    OB2PORTRANGESPEC=xMA-NET:20100-20150

    now all data streams for backup/restore use this port range (DA->MA)
  4. now configure your tmg to allow tcp 5555 from Cell Server to DMZ Server and TCP 20100-20150 from DMZ Server to Media-Agent (Cell Server??)
  5. test if you can telnet dmzserver.domain.local 5555 from cell server. if you can't connect. something with name resolution or tmg firewallrules are wrong.
  6. if you setup everythin right, it should work now

 

Regards

↧

Re: Backup client on DMZ

November 22, 2013, 4:00 am
$
0
0
try adding OB2INETTIMEOUT=60 to the dmz server in omnirc file
↧
Search

omnidbcheck doesnt complete HPDP 6.20

November 22, 2013, 5:05 am
$
0
0

The IDB Daily task has stopped completing recently, when i open a command prompt and run

 

omnidbcheck -quick

 

it just never seems to finish (dp40 folder is 17GB) and the CPU on the machine is almost 100% idle, what can i do to find out what the problem is as it seems to create a Check_core.txt file in C:\Program Files\OmniBack\log\server but its 0 bytes in size...

 

any advice would be greatefully appreciated 

↧
↧

Re: how to configure multistreaming on backup to disk

November 22, 2013, 5:22 am
$
0
0

I have over 100 systems.  in my environment that would mean 2-300 virtual drives.  seems a little high.  i suppose there is no advantage in exceeding the number of virtual drives over what the load balancing max limit is set to?

thanks

↧

Re: DP 8: Error bkp IDB - Putting the Internal Database into the backup mode failed

November 22, 2013, 6:06 am
$
0
0

Thanks for your opinion, Bob_Clark, will open case in support HP.

Hilton Junior

↧

Re: how to configure multistreaming on backup to disk

November 22, 2013, 6:16 am
$
0
0

Does each system have their own VTL?  That may not always be the best way to do it, especially if the data is similar to several other servers.

 

Let's say you have 10 Windows fileservers and 10 Linux application servers.  Would it make sense to backup all systems to one location (VTL)?  No, not really.  Your deduplication will suffer therefore the data will take up more space on disk.  The solution is to create TWO VTL's, not 20.  The 10 Windows servers will backup to one VTL, and the Linux servers will backup to the other.

 

So, just think about that before you start going crazy with VTL creation.

 

But I digress, how to configure multistreaming.

 

For me, the number of backup streams is generally determined by the 'Backup Object Summary' page.  If you're backing up an entire system and only one item is listed under that tab, that means ONE stream will backup the entire server.  This isn't always the best.  Let's say you have a server with a C, D, and E drive.  Here's how I would configure it.

 

1. Create a Blank Filesystem Backup

2. Find the server, then check if off

3. Expand it and UNCHECK the D and E drive

4. Right-click on the server, choose Copy

5. Expand the duplicate server, check D

6. Copy the server again.

7. Expand the third duplicate server, check E.

8. Next, then change the backup destination and change the Max value to 3

 

As you progress forward, you'll see three rows in the Backup Object Summary page.  That means there will be three streams.  This practice has worked really well for me, and I've been more granular too.  Let's say you know there are four directories on the root of D that take up a lot of space.  You could follow the same practice and only choose that directory, then copy the server, choose the next folder, and so on.

 

Another thing of importance is the Order field in the summary page.  If you want the largest directories/drives backed up first, change the Order by right-clicking the line then Move up or Move down.  That feature isn't advertised well and it's difficult to find too.

 

Let us know how you make out!

↧

Re: omnidbcheck doesnt complete HPDP 6.20

November 22, 2013, 7:10 am
$
0
0

First, you need to get your IDB backed up.  In the Backup specification for your Internal Database, go to the Backup Object Summary tab.  Right-click in the [Database] object, and pick Properties, or just click the Properties button

 

Go to the Internal Database tab, and un-select "Check Internal Database". Click OK and Apply.  This will enable you to continue backing up the IDB while you get teh 'omnidbcheck' problem sorted out

 

You can take the parts of the 'omnidbcheck -quick' or 'omnidbcheck -extended' and run them individually

 

          omnidbcheck -dc

          omnidbcheck -bf

          omnidbcheck -filenames

          omnidbcheck -core

 

to see if there is any corruption or inconsistency being detected

 

I think if you waiting long enough, several hours perhaps, the 'core' part would come back with something, since the IDB is 17GB.  But, how long since you did a purge of your IDB?

 

         omnidbutil -purge -filenames -force

         omnidbutil -writedb ..........

         omnidbutil -readdb ..........

 

If it has been quite a long time, or, if you never remember doing it at all, I think it needs to be done

 

 

↧
↧

Re: Overall Deduplication Ratio

November 22, 2013, 8:21 am
$
0
0

Where you're seeing the reports, on server or in other machine?

 

I had this issue just after update my environment.

 

When I was seeing the reports on the server it did not happen. Only happen in the client side. The problem was solved applying the updates in the user interface on the machine  where I usually see the reports.

 

↧

Re: how to configure multistreaming on backup to disk

November 22, 2013, 8:32 am
↧
Search

Re: how to configure multistreaming on backup to disk

November 22, 2013, 8:40 am
$
0
0

nearly all windows systems here.  one backup server with one VTL.

great advice on breaking up the system drives.  thanks.

i just found the option to change the backup order the other day too.  Doesn't look like there's an easy way to get the one sitting @ 89 up to #1.  you can only move it up one step at a time.  <sigh>

thanks .

 

↧

Re: how to configure multistreaming on backup to disk

November 22, 2013, 8:47 am
$
0
0
So you have one backup spec for ALL windows servers? How many drives on the VTL?
↧
↧

Re: how to configure multistreaming on backup to disk

November 22, 2013, 8:57 am
$
0
0
rgb99 wrote:
So you have one backup spec for ALL windows servers? How many drives on the VTL?

Well, you could put all Windows servers in a single backup spec but you risk maxing out the media servers network bandwidth resulting in a very long session.

 

My preference is to multi-stream to the max in each spec taking into consideration bandwidth, CPU load on the client (each stream starts a new VBDA on the client), speed of disks and other tasks that the client may be performing during the backup.

 

This is very much a trial and error scenario which depends upon your particular environment.

↧

Re: Overall Deduplication Ratio

November 22, 2013, 10:42 am
$
0
0

I found a lab case where a similiar problem was fixed with a test binary.  If you are already using DP 7.03, you would need to open a case with the response center, and refer to

 

 

QCIM2A46109 QCIM2A45607_TM1 Results in incorrect size reported for "Mbytes Total"

 

↧

Re: Problems with Oracle/SAP Backup DP 6.11

November 23, 2013, 3:44 pm
$
0
0

Hello,

 

Have this been resolved?  I am having the same error.  

 

Thanks,

 

Tony

↧
Viewing all 10494 articles
Browse latest View live
Search
<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>