Jump to content

best approach to multi-server backup


Recommended Posts

Hello everyone!

Just wondering if it better backup mulitiple servers (from rertospect server) by accessing drives/share on other servers vs. having clients installed on each additional server?

I have noticed that my tape times out after even 1 hrs of trying to backup some shares....I am guessing it is due to the fact that some files are in use on the drive/share.?!( I dont have the licence for Open Files...would this solve the issue?

so which apporach would be better?

eg.

A

[Rertospect server]

|

|

|x:drive

|[file server 1]

|

|

|y:drive

|[file server 2]

 

 

vs.

B

[Rertospect server]

|

|

|client on

|[file server 1]

|

|

|client on

|[file server 2]

 

 

 

Many thanx for your time!

All the best!

Link to comment
Share on other sites

Hi Nate!

Thanks for the respond! Appologies for not writing back sooner!

 

I am a bit confused...why is there an option to use "My Network Places" under volumes selector if is not suggested?

My thinking was that "retrospect clients" would add an overhead to other file servers (and network) and being that shares are there a I used 'volumes' apporach.?!

 

Could this be the reason to my dreaded Media Timeouts?

I am getting really frustrated with this and not sure where to turn....

 

I am using Folder Redirection of "my documents" and files are redirected to the FileServer1 to a share...this share (from Retrospect's perspective) is the one defined as a Volume 'Users'

 

- 4/27/2005 6:34:34 AM: Copying Users on FileServer1

Media request timeout after waiting 1:00:00

4/27/2005 7:41:55 AM: Execution incomplete

Remaining: 5586 files, 3.0 GB

Completed: 1822 files, 532.7 MB

Performance: 78.1 MB/minute

Duration: 01:07:20 (01:00:31 idle/loading/preparing)

 

 

I guess the ultimate question would be... does the media timeout happen due to the fact I am using volumes apporach vs. clients or is there something else?

Strange thing is that the rest of the volumes do NOT timeout and are backed up just fine...network performance is not an issue....files are not in use.... and yet the back up of 20-25 GB takes about 9 hrs!!!!!!

HEEEEELP!

Many thanks for sharing!

All the best!

Link to comment
Share on other sites

Hi

 

The client adds minimal overhead. You won't notice at all and you won't have to change your volume login settings everytime a domain password changes.

 

Take a look at your script settings and backup set properties. I suspect Retrospect is looking for media that isn't present. Have you considered turning on the email notification in Retrospect? This will tell you exactly which media Retrospect is looking for.

 

Thanks

Nate

Link to comment
Share on other sites

Hi Natew!

 

Thank you for the help so far, I really apprechiate your time put into these forums!

 

It sounds like I really should use clients from other server's perspectives so will give it a try.

 

 

 

I checked the scripts and everything seems fine. I have used your advice to switch from normal to recycle backups and tapes are found without problems.

 

 

 

If I run the script manually it backups everything fine, the only problem that is left is the issue 'time out' when trying to access my last volume mentioned above. Again, it only times out when I run scheduled backup. If I browse the volume the response time is great....3.5 GB worth of files are listed in a matter of 4-5 seconds so I know that speed of network(drive/volume) is not an issue.

 

In hand with the 'time issue', can it be the process of "comparing' that slows down Retrospect?

 

Here is sample:

 

--------------------------------------------------------------------------------------------

 

- 4/28/2005 4:03:24 AM: Copying Documents and Settings on FileServer3

 

4/28/2005 4:55:49 AM: Snapshot stored, 53.0 MB

 

4/28/2005 4:55:59 AM: Comparing Documents and Settings on FileServer3

 

4/28/2005 6:40:06 AM: Execution completed successfully

 

Completed: 96860 files, 2.9 GB

 

Performance: 37.7 MB/minute (57.3 copy, 28.1 compare)

 

Duration: 02:36:41 (00:02:20 idle/loading/preparing)

 

--------------------------------------------------------------------------------------------

 

 

 

Should I turn of "verification" to speed things up or is this not a good idea?

 

 

 

Perhaps the solution is in using the clients thoughout.

 

Many thanks for all the help!

 

All the best!

Link to comment
Share on other sites

Hi

 

Never turn off verification. You will live to regret it.

 

Media time outs will only happen when Retrospect is looking for media to back up to. They won't happen when a network volume is not available. Are you saying that you can manually run the script from the "run" menu in Retrospect and it will run normally?

 

Thanks

Nate

Link to comment
Share on other sites

Hi Nate!

Thats is exactly it, I even confirm that scipt is ready and that tape is found a few hrs before scheduled execution.

 

here is the example of scheduled run from the 1st post:

- 4/27/2005 6:34:34 AM: Copying Users on FileServer1

Media request timeout after waiting 1:00:00

4/27/2005 7:41:55 AM: Execution incomplete

Remaining: 5586 files, 3.0 GB

Completed: 1822 files, 532.7 MB

Performance: 78.1 MB/minute

Duration: 01:07:20 (01:00:31 idle/loading/preparing)

 

 

I really find it hard to belive that it takes (and fails) more than 1 hrs to backup 532.7 MB....how come it can backup 8-9 volumes before it gets to the last one? my tape doesnt exaclty jump out of the drive in the middle of the back smile.gif

 

again this volume is the last listed in my scipt if it makes any difference.

In hand, why would the snapshots need to be 'comapred' if my backup is set to recycle..? shouldn't it just overwrite the content of the tape no matter what is on it?

I will try using email notification to see if get any more details....

Thanks for all the help!

All the best!

Link to comment
Share on other sites

Hi again!

Email notifications did not reveal much other than what logs show...

 

From Retrospect: Script "2-FRI_NetworkFiles" failed during automatic execution, error -1204 (user didn't respond). Please launch Retrospect and check the log for details.

 

..still no love on this timeout issue.... frown.gif

Link to comment
Share on other sites

frown.gif

it this the point where I sould consider a new product? I am still looking for some solution to these -1204 errors...... it seems that RS can't really get the job done.

The stats below are just horrible...

a backup of 30 GB takes more than a full working day?

 

5/3/2005 7:43:11 AM: Execution incomplete

Total performance: 146.1 MB/minute <------- if this is correct then backup should take no longer 3 hrs

Total duration: 08:43:03 (01:12:57 idle/loading/preparing)

 

 

Anyone has suggestion or has found the soution to tape timeouts? ...can it be that NTbackup can outperform RS?

Much oblidged for any suggestions...

All the best!

Link to comment
Share on other sites

  • 3 weeks later...

Hi

 

Quote:

I really find it hard to belive that it takes (and fails) more than 1 hrs to backup 532.7 MB....how come it can backup 8-9 volumes before it gets to the last one?

 


 

You are right - it doesn't take that long. Most of that hour Retrospect is waiting for the user to insert a tape. This is most likely because the first tape is full. If you go to the properties of your backup set does it say "appending requires another member" on the main summary window? How much space does it say is used on the first tape.

 

Thanks

nate

Link to comment
Share on other sites

  • 1 month later...

Hi Nate!

Sorry for not getting back sooner, I though that thread has died.

I have switched to using clients on the other servers and performance seems to have increased. I still do have the same issue of timeouts on one of the volumes...there are no other errors other than timeout issue.

I backup about 32 GB on the 40 GB tape so there is enough space.

The volume in question is the volume where 'my docs' folder is redirected for all users so my best reasoning for timeouts would be that some files are in use while backup is running...hence the behavior showing that sometimes timeouts happen after successful backup of [2.3GB of 4 GB] or [1.5 GB of 4 GB] etc.

Bigger isssue than 'not all files backed up' is that backup stops after this failiure/timeout on this volume and doesn't continue backing up other volumes...

 

Am I missing 'an option' that will ignore these 'in use files' and continue with its job.

Many thanks for following this thread!

All the best!

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...