cancel
Showing results for 
Search instead for 
Did you mean: 

Secondary Node get Panic

vinods
Level 5
Partner

Hi All,

We are using VVR 5.1 SP2 on Windows2008 R2 for two node replication of file server data. Intially we had set "UDP" protocol for replication but found that replication was too slow on 45 Mbps WAN link we also tried changing packet size but all in vain. So we decided to change the protocol from "UDP" to "TCP" after changing the protocol replication speed increase but within three to four mins secondary node gets panic automatically. After through inspection we found that node gets panic only when we specify bandwidth above 10Mbps  and works fine if bandwidth is < = 10 Mbps.

 

Note:- Secondary Node doesnt panic if Protocal is "UDP" and bandwidth is > 10 Mbps or "Maximum Available" but problem with "UDP" protocol is replication is too slow.

 

Thanks & Regards,

Vinod Shinde

1 ACCEPTED SOLUTION

Accepted Solutions

Wally_Heim
Level 6
Employee

Hi Vinod,

Just thinking about it, I would like to add a switch for extended logging to the second log to collect on the primary.  Please use this instead of what I have above.

On primary:

2. vxrlink -i 5 -e stats <rlink name> >c:\pristats.txt

The rest of the steps are the same.

Thanks,

Wally

View solution in original post

4 REPLIES 4

Wally_Heim
Level 6
Employee

Hi Vinod,

I would recommend that you open a support case for both of your issues.  We can tune VVR for better throughput with UDP and we need to investigate the reason for the BSOD on the secondary server.

 

Issue #1:  BSOD on secondary when replicating in TCP mode with Bandwidth limit > 10 mbps.

Action needed for further troubleshooting:  If created we will need to memory dump on the secondary and a set of vxexplorer logs.  If a memory dump was not create please configure the server to capture a kernel level memory dump and reproduce the issue to create the memory dump.

 

Issue #2:  Replication performance issue when replicating in UDP mode.

Action needed for further troubleshooting:  With replication mode set back to UDP please collect the following logs:

 

On primary:

1. vxmemstat -i 5 >c:\primem.txt

2. vxrlink -i 5 stats <rlink name> >c:\pristats.txt

3. vxrlink -i 5 status <rlink name> >c:\pristatus.txt

4. a set of vxexplorer logs

5. vxprint -VPl >c:\privxprint.txt

 

On secondary:

1. vxmemstat -i 5 >c:\secmem.txt

2. a set of vxexplorer logs

3. vxprint -VPl >c:\secvxprint.txt

 

Thanks,

Wally

Wally_Heim
Level 6
Employee

Hi Vinod,

Just thinking about it, I would like to add a switch for extended logging to the second log to collect on the primary.  Please use this instead of what I have above.

On primary:

2. vxrlink -i 5 -e stats <rlink name> >c:\pristats.txt

The rest of the steps are the same.

Thanks,

Wally

vinods
Level 5
Partner

Hi Wally ,

 

Sorry for the delay pls find the attached logs of the  above  command, Due to size i am not uploded vxexplorer logs.

Wally_Heim
Level 6
Employee

Hi vinods,

 

As mentioned, you need to open a Symantec Technical Support case for this issue and provide Support with the logs mentioned.

 

Thanks,

Wally