cancel
Showing results for 
Search instead for 
Did you mean: 

Error 41 Network Connection Timed Out

dyerja
Level 3
Have a server, with the following setups:

NetBackup 5.1 MP2 on Windows 2003 Server
Server is the Master Server and Media Server, and trying to back up the master server's E: partition via a  FlashBackup-Windows policy \\.\E:  - 243 GB partition

This initially started with the backups failing with an Error Code 41 Network Connection Timed Out.  Backups prior to 4/16 (when this issue initially started) were fine.

Changed the timeouts up to their maximum value and now the full backup will complete after about 10 - 12 hours of nonactivity it will write the full backup to tape in those last hour and a half or so and then complete successfully.  This seems to be a bit too long for the backup to complete.

The differential backups that run during the week however, run for about 12 hours or so and then error out with the Error Code 41 Network Connection Timed Out.

During both types of backups the bpbkar32.exe process continually pegs at about 25% CPU time.  Tried setting the policy to "Allow Multiple Data Streams" but not sure how this will really help if the backup is being done via FlashBackup-Windows.

Now, updating to a later 5.1 MP isn't an option at this point in time (however, if there is a known issue with 5.1 MP2 with something like this, I'd appreciate it if someone would let me know), right now I'm grasping for straws at this point.

Research into FlashBackup has led to info on it about how it performs its backups and how it parses through the filesystem to create the snapshot, so I'm also concerned if this is the wrong way to go about backing up this system.

Please let me know,  Thanks!

Joseph Dyer


Message Edited by dyerja on 05-01-2008 07:55 AM
9 REPLIES 9

NY2AZ96
Level 3
Employee Accredited

For a FlashBackup client, this can happen if the file system being backed up is very large and has a very large number of files. It can also occur if a large number of concurrent data streams are active at the same time. The corrective action is to add CLIENT_READ_TIMEOUT to the /usr/openv/netbackup/bp.conf file and set it to increase the timeout interval.

dyerja
Level 3
Sorry, but no. Smiley Wink

Considering that's right out of the Veritas Troubleshooting guide, I doubt that's our instant fix to this.

Ron_Cohn
Level 6
Did you by any chance add or create "exclusion" entries for this server?  Also, does the drive have a large number of small files?

dyerja
Level 3
There are exclusion entries listed under the host properties of the system yes.

And yeah, there are a large number of smaller files, but I have another system that's essentially the same, and it works fine there.

Omar_Villa
Level 6
Employee
how are the buffers configured? you can have a bottleneck there, also, check the bpcd and bptm logs and grep for "time out" or "delay" there you will see where the issue can be.
 
 
regards

dyerja
Level 3
Buffers are set to the defaults.

I do not see anything in the bptm or the bpcd logs pertaining to delays or timeouts.

I have noticed that the bpbkar log that it hangs at a line

11:36:32.625 PM: [3808.2976] <2> tar_base::V_vTarMsgW: INF -      clientname imgsv01, buid imgsv01_1210044946, filnum 0

and this is the line it typically hangs at when the job starts.  In the admin console, under Activity Monitor the number of files written hits 4500, number of kilobytes written is 0.  It also says it has begun writing (but when you look at the tape drive, it's not doing anything)

Thanks!

Ron_Cohn
Level 6
The reason for my query about the exclusion list is this:  I have 2 policies.  A Standard Backup and a FlashBackup.  In the client configuration, listed under All Policies->All Schedules, I listed my E:\ drive as being exlcuded.  On that same screen, you can override that exclusion by saying that exclusion does not apply to the following policies.
 
What it does, NetBackup builds an exclusion table, containing *each* filename on my E:\ when running the *exception* policy.  Make sure you are not following this scenario.

dyerja
Level 3
Ron,

Thanks, that clarifies that a lot.  lol.

What we're doing is an All Policies -> All Schedules exclusion list.  We don't have any exceptions listed, and the only things we're really excluding is what appears to be default NetBackup exclude lists (veritas bin files, etc. pagefile.sys (on c:\), and mdf, ldf files) - we use SQL but we back up the SQL backups that are created from SQL instead of having an actual SQL agent for it.

I really appreciate all the help here guys.  This is a weird issue and what really boggles me on it is having another one that works perfectly fine.

Thanks again!

Joe

dyerja
Level 3
Anyone have any other assistance on this?  Just wondering if anyone else might have any ideas for this issue.

Thanks!