Skip to main content

Hi!

We recently had an issue that led to the Exchange admins reinstalling the Exchange hosts. Once they were installed, i installed the commvault agent again, but i’m running into an issue that i’ve been unable to solve so far.

It’s a small exchange environment with only 2 virtual exchange servers.

What i do know is that during the re-installation of the exchange servers, according to the exchange admin, one of the two were originally installed with one name, before they were renamed, and then exchange was installed.. 

 

However in Commvault i’m running into an issue where if i do a “Check Readiness”, both the servers says they are ready, while the DAG-client complains. 

Further research into logfiles led me to the cvfwd.log files, where i see references to both the current name, and the name one of the servers temporarily had, before exchange was installed/configured, and this seems to cause the tunnels between the “components” to terminate, causing communication issues.

 

Example: 

18028 9e64 02/15 14:48:52 OT:00007 ######## :Exchangeserver20] This is a control tunnel. Deleting all extra tunnels.

18028 9e64 02/15 14:48:52 ######## CN:00015 :local->Exchangeserver20:8400] ERROR: cvfwd_patch_client(): Failed to find tunnel to Exchangeserver20
18028 9e64 02/15 14:48:52 DT:00014 ######## ERROR: cvfwd_iot_wait(): Socket READ failed. Got READ error on DYNAMIC tunnel via (A.B.C.38, A.B.C.38) on port=0, fd=1032: The specified network name is no longer available.
18028 9e64 02/15 14:48:52 DT:00015 ######## ERROR: cvfwd_iot_wait(): Socket READ failed. Got READ error on DYNAMIC tunnel via (A.B.C.38, A.B.C.38) on port=0, fd=1036: The specified network name is no longer available.

 

Now, Exchangeserver20 does not exist, it’s a temporary name that was used during the re-install, and i’m at a loss as to how/where Commvault is picking that up from  ?

 

The actual error i get when trying to run backups is the following: 

Error Code: o19:1597] Description: Failed to start phase iDiscover] on Exchangeserver11] due to network error eRemote system Exchangeserver11.Domain*Exchangeserver*8400]. Communications Service returned error when executing remote command wexDBStartBackup.exe] 

 

 

I’ve attached a sanetized version of the cvfwd.log file after a restart of the services.

 

Hey @Bjorn M 

I think the network config is using the name Exchangeserver20 for the local connections to Exchangeserver10, it’s detecting the name somewhere and it’s using it:

 

18028 9e64 02/15 14:48:52 OT:00007 ######## #<Exchangeserver20>] Initialized new ON_DEMAND control tunnel from "<Exchangeserver10>" to "<Exchangeserver20>" via (<A.B.C.>38, <A.B.C.>38)
18028 9e64 02/15 14:48:52 OT:00008 ######## #<Exchangeserver20>] Gracefully stopping tunnel (ON_DEMAND control tunnel from "<Exchangeserver10>" to "<Exchangeserver20>" via (169.254.1.223, 169.254.1.223)) due to duplicate elimination
18028 9e64 02/15 14:48:52 OT:00007 ######## #<Exchangeserver20>] ERROR: do_recv_cname(): The reported client name <Exchangeserver10> doesn't match the one in the local config: <Exchangeserver20>. Keeping <Exchangeserver20>.
18028 9e64 02/15 14:48:52 OT:00008 ######## #<Exchangeserver20>] ERROR: do_recv_cname(): The reported client name <Exchangeserver10> doesn't match the one in the local config: <Exchangeserver20>. Keeping <Exchangeserver20>.
18028 9e64 02/15 14:48:52 DT:00011 ######## ERROR: cvfwd_reset_tunnel(): Ungraceful termination of DYNAMIC control tunnel via (169.254.1.223, 169.254.1.223)

Can you check in the Base directory on Exchangeserver10, you’ll see 3 txt files - FWConfig, FWConfigInstall, and FWConfigLocal.

 

Can you see if theres any reference to Exchangeserver20 here? If not check in the Windows registry under HKEY_LOCAL_MACHINE\SOFTWARE\CommVault Systems\Galaxy\Instance001. Do a find here for Exchangeserver20 and see if anything comes up under the Commvault hive (don’t make any changes to the reg) just let me know what you find.

I would suggest raising a support incident as a remote session is likely required to rectify this. 

@Matt Medvedeff 

The 3 files are empty (well 1 contains some config-examples txt that’s commented out, but no references to any servers)

And there is nothing with Exchangeserver20 in the registry.

@Chris Hollis 

That seems to be the direction we’re heading unless we find some solution :-)

 

 

Problem seems to be fixed, and it was just an incredibly stupid thing….Exchange agent/component had not been installed again after one of the multiple attempts that was done early on in the process of fixing the issue, where it had been uninstalled.

 

 

Reply


Terms & ConditionsCookie settings