Client Agent Reconnection Timeout

[ross7]

I'm trying to backup a database through the client agent. I'm using Brightstor r11.1 and client agent is r11.

The problem is the backup always comes back as failed with a reconnection timeout error. I appear to be getting the most of the data but the job always shows failed.

I tried increasing the send receive timeout options but with no luck. Below is a sample of the agent log. Any ideas?

03/06/05 01:01:26 =============== Start Backup Job ====================
03/06/05 01:01:26 Local Options for NOTTINGHAM_BDC: control ID - 2400,2670.
03/06/05 01:01:26 --------------------------------------------------
03/06/05 01:01:26 DataTransferBuffer = 32768Kb.
03/06/05 01:01:26 SendTimeOut = 2400sec [40min].
03/06/05 01:01:26 IOBuffer = 512Kb.
03/06/05 01:01:26 TLIBuffer = 512Kb.
03/06/05 01:01:26 NoB = 1Kb.
03/06/05 01:01:26 AIO enabled, 4 buffers, Size = 1024Kb.
03/06/05 01:01:26 Reconnection enabled
03/06/05 01:01:26 SetTCPnoDelay enabled.
03/06/05 01:01:26 --------------------------------------------------
03/06/05 01:01:26 [AADR] - Server support disabled or not present.
03/06/05 01:01:26 Start Backup Session #1.
03/06/05 01:01:26 AgentInfo = V: 11.0 B: 2670, tB: 6e0a, mbf: 32768
03/06/05 01:01:26 Tape Name = SUNDAY
03/06/05 01:01:26 (sesType, sesMethod) = (42, 11)
03/06/05 01:01:26 Common Path = C:\PARCSOFT\SERVER\V3\Htepsi\db
03/06/05 01:01:26 ASDB_BEGIN_SESSION OK.
03/06/05 01:01:26 TP_BEGIN_SESSION OK.
03/06/05 01:01:26 Backup Estimation started
03/06/05 01:01:26 Backup Estimation done.
03/06/05 01:01:26 TP_RESUME_DAEMON OK.
03/06/05 01:01:26 TP_GET_FREE_BUFFER OK.
03/06/05 01:01:26 [SD] Start new SD...
03/06/05 01:01:26 [INITIALIZE] MB: 33554432, MPS: 2097152, CS: 20132659
03/06/05 01:01:26 SendPacketSize = 1088Kb.
03/06/05 01:01:26 Socket send buffer size: 3342336
03/06/05 01:01:26 iCount = 8, timeout = 300000, pCD SendTimeout = 2400000
03/06/05 01:01:26 BOS OK.
03/06/05 01:02:53 Send EOS OK
03/06/05 01:02:53 Backup operation done.
03/06/05 01:02:53 [WD] Initiate end procedure...
03/06/05 01:02:53 Watch dog daemon ended successfully...
03/06/05 01:02:53 [SD] Close procedure started...
03/06/05 01:02:54 [SD] Wait until the data buffer became empty. RC=258
03/06/05 01:02:55 [SD] Wait until the data buffer became empty. RC=258
03/06/05 01:02:56 [SD] Wait until the data buffer became empty. RC=258
03/06/05 01:02:57 [SD] Wait until the data buffer became empty. RC=258
03/06/05 01:02:58 [SD] Wait until the data buffer became empty. RC=258
03/06/05 01:02:59 [SD] Wait until the data buffer became empty. RC=258
03/06/05 01:03:00 [SD] Wait until the data buffer became empty. RC=258
03/06/05 01:03:01 [SD] Wait until the data buffer became empty. RC=258
03/06/05 01:03:02 [SD] Wait until the data buffer became empty. RC=258
03/06/05 01:03:02 [SD] Initiate end...!
03/06/05 01:03:02 [SD] Cleanup...
03/06/05 01:03:02 [SD] Wait until the data buffer became empty. RC=0
03/06/05 01:03:02 [SD] End...
03/06/05 01:43:03 Exit procedure started...
03/06/05 01:43:03 1 folders 2 file(s) 276648 KB written to FRIDAY @ 190417 KB/min
03/06/05 01:43:03 [SD] SDT is closed.
03/06/05 01:43:03 Backup operation failed.
03/06/05 01:43:03 Exit procedure done.

 

[davidmichel]

Strange. Data gets backed up ok, but then there is something going on in closing up the job that does not get done. Not sure if this is at the Server or client end.

If it has not already been done but the Job Engine in debug mode. Also after there are no jobs running check the ARCserve\Log directory for files. When there are no jobs running it should be empty. The existance of files is an indication that the Database is not getting updated.

 

[ross7]

Thanks for the reply David. I've set the job engine to debug and I'll see what it shows tomorrow. I think the problem must lie with the client agent as I also backup over 20 other remote sites with the same configuration without issue.

I had a look in the Arcserve\Log file as you suggested and there are a few files in there when no job is running.These have the extentions .log_dix, .log_lix and .log_jix.

 

[sanbe03]

hi,
there is a document on the CA site :
TEC266855 Troubleshooting E8522 and E3392 (Failed to receive data from the client agent) errors
Could be helpful if you have network traffic problem, because there are registry key to change the timeout.