windows 2003 event id 1053 netbackup 6.5 console error

[olivieri_tizian]

Hi

I have installed windows 2003 ST ED with SP2 and patch, with Netbackup 6.5 updated from 6.0 MP4.

I receive this message about 3 o 4 days:

the problem seems about tcp copnnectio opened, are there more than 100 connection TCP state in CLOSE_WAIT on the NetBackup Server  !!

TCP    itasmi09:5952          itasmi09:vnetd         CLOSE_WAIT      1208
  [bmrbd.exe]

  TCP    itasmi09:5963          itasmi09:vnetd         CLOSE_WAIT      1208
  [bmrbd.exe]

  TCP    itasmi09:5977          itasmi09:vnetd         CLOSE_WAIT      1208
  [bmrbd.exe]

  TCP    itasmi09:5979          itasmi09:vnetd         CLOSE_WAIT      1208
  [bmrbd.exe]

  TCP    itasmi09:5981          itasmi09:vnetd         CLOSE_WAIT      1208
  [bmrbd.exe]

  TCP    itasmi09:5989          itasmi09:vnetd         CLOSE_WAIT      1208
  [bmrbd.exe]

  TCP    itasmi09:5991          itasmi09:vnetd         CLOSE_WAIT      1208
  [bmrbd.exe]

  TCP    itasmi09:5992          itasmi09:vnetd         CLOSE_WAIT      1208
  [bmrbd.exe]

  TCP    itasmi09:5994          itasmi09:vnetd         CLOSE_WAIT      1208
  [bmrbd.exe]

I've worked about this problel with Microsoft Support and they ave suggested me to add these keys in regedit :

- For Backup server

1. We can help by telling TCP to close ports that have been inactive for 30 seconds by changing the following key:[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters]

"TcpTimedWaitDelay"=dword:0000001e

"MaxUserPort"=dword:0000fffe

 MaxUserPort info:

The problem is partial solved, the server seems works about last 4 days, but with netstat command I can see again more TCP connection in CLOSE_WAIT

Please, Do you have an idea to sollve this issue ?

Thanks

[14ALL]

I have the same problem.  Veritas NetBackup v6.5 running on Windows 2003 Standard SP2

NetStat -a shows the following:

TCP    Server1:2902     Server1:domain.com:vnetd    CLOSE_WAIT

TCP    Server1:2904     Server1:domain.com:vnetd    CLOSE_WAIT

TCP    Server1:2906     Server1:domain.com:vnetd    CLOSE_WAIT

Ports in "CLOSE_WAIT" continue to multiply until the server runs completely out of ports at which time no further backups can run until the server is rebooted.

We tried adding/modifying Registry settings, but no change.  One of our Admins has been working with Symantec Tech Support for a week with no resolution. 

What is causing this and how do we fix it?

[14ALL]

Still no solution.  Any suggestions?

[14ALL]

Update: No longer having issue of "vnetd CLOSE_WAIT" consuming all server ports.  
1.  Upgraded to NetBackup 6.5.4. 
2.  Uninstalled 6.5.4 updated due to old client dependencies.
3.  Upgraded to v6.5.2 which can backup older AIX Servers running client v5.
4.  Problem with vnetd consuming ports eliminated.

Background: During initial build of our single master server, we experienced difficulty with NIC Teaming and designating ports for admin console access.  In the process we ended up with a phantom nic (sounds scary I know).  To remedy the phantom problem, Dell send us a vbscript with over 2,000 lines of code in it.  The script didn't fix the problem and a call to Intel provided the simple solution (minus thousands of lines of code).  Since I have not seen a flood of posts on this issue and working with Symantec Support did not produce resolution, I suspect that the root cause may have had something to do with the initial build difficulties.

If having problems with unknown devices, consider Microsoft Doc http://support.microsoft.com/kb/315539 

====================================
Phantom NIC Solution (for anyone interested)
Intel had me download two utilities (PROUnins.zip and P3uninst.exe) then perform the following steps in sequence.

Launch Device Manager. 
Delete all 4 Intel Pro 1000 PT Nics (two 2-Port nics installed)
Control Panel.  Uninstall Intel Pro 1000 software
Run ProsetUninstallerNT.exe from the PROUnins.zip Intel web site download.
Reboot Server
Run P3uninst.exe to remove Phantom NIC (also downloaded from Intel’s web site)
Install Intel driver v12.3 and configure new Team.
 
They informed me that there is a known issue when doing remote desktop into a server and trying to view the teaming tab.  It won’t work by default, however, the teaming tab should still be present when at the server console.

Instructions for fixing this problem are found at http://www.intel.com/support/network/sb/cs-020166.htm

[Zahid_Haseeb]

 HI olivieri what the impact comming on your netbackup because of this ? i mean any slow speed, or any backup/restore fails ?

[14ALL]

To answer Zahid Haseeb's question, once all ports got used up, the backup server would stop working.  You couldn't even launch NetBackup Admin Console and would have to do a hard boot to get the server working again.  Until we performed the above solution (Update: No longer having), we had a script doing nightly reboots to clear ports.  We discovered the solution by upgrading, running into client issues and then downgrading.  It is possible that a simple reinstall of 6.5.2 may have also fixed our problem.