05-15-2014 02:17 AM
hi all!
i have a problem backing up a netapp filer via NDMP, but only sometimes
setup: Single Masterserver/Mediaserver with W2k8R2 and Netbackup 7.5.0.5, LTO-Library and Netapp-3020 Filer
the NDMP dump itself is successful, both times, but thereafter, something goes wrong
i attach 4 logs, one the log of the job seen in the GUI, the other one being bptm reduced to the job in question & the final moments (some info had to be replaced by policy, sry), once for the failed job, once for the successful one - to be specific, its actually the second try of the same job
trouble begins at line 56 of the bptm*_end.txt where on one attempt the connection succeeds, and fails on the other..
it seems to be a connection issue, which is strange though, since none of the listed ports is explicitly allowed by the FW inbetween the two networks (10.32.238 and 10.61.192) except the ones specified by symantec (10000, 1556, 13720, 13724, 13782, 13783)
it also seems that the problem occurs on large backups more frequently than on smaller ones
should further info be required let me know
if someone could help me to understand the issue, and if its really necessary to open a wide range of ports (although i doubt that)
thanks for your time
rob
p.s. i checked all the technotes about NDMP, Error 23 and the forums
also tried setting NDMP_PROGRESS_TIMEOUT and various other (but NOT NDMP_MOVER_CLIENT_DISABLE)
05-15-2014 05:27 AM
05-15-2014 05:46 AM
hi!
yes, i noticed, but i do not want to use this, since it works on some occasions, and i would like to find the cause, before i use this option
if my backups would fail all the time, i would've tested this option already
but thanks for your comment :)
05-18-2014 06:38 AM
Good port was 1863, bad one was 2594
I usually have customer's doublecheck that ports 1025-5000 are open for the data connection server that we open.
05-18-2014 07:38 AM
Check out the following:
http://www.symantec.com/docs/TECH205899
That apepars to show the same issue you're experiencing.
From your logs:
17:53:59.370 [4856.3416] <2> NdmpMoverSession[0]: ndmp_mover_connect failed, status = NDMP_CONNECT_ERR
17:53:59.370 [4856.3416] <2> NdmpMoverSession[0]: MoverConnect method failed. Reinitializing and attempting to use MoverListen method.
17:53:59.604 [4856.3416] <2> NdmpMoverSession[0]: ndmp_mover_set_window(0, 18446744073709551615)
17:54:00.025 [4856.3416] <2> NdmpMoverSession[0]: ndmp_mover_listen failed, status = NDMP_CONN_TYPE_UNCONFIGURED_ERR
17:54:00.025 [4856.3416] <16> NdmpMoverSession[0]: ERROR Start failed
17:54:00.025 [4856.3416] <16> check_and_process_mover_tasks: NDMP Mover Client Setup failed
17:54:00.243 [4856.3416] <2> NdmpMoverSession[0]: Shutdown complete
17:54:00.243 [4856.3416] <2> write_data_tir: ndmp_task = 1
17:54:00.243 [4856.3416] <2> write_data_tir: status_to_return = -1
17:54:00.243 [4856.3416] <2> check_error_history: just tpunmount: called from bptm line 21443, EXIT_Status = 23
17:54:00.243 [4856.3416] <2> io_close: closing E:\Program Files\Veritas
If that doesn't resolve the issue - check out the following as it has great networking troubleshooting tooks.
05-18-2014 11:58 PM
hi deb!
saw that one too, but again, the backup is not failing all the time, so i do not think its a resolver issue, as the name is correctly resolved in the logs as well
and, besides that, i have two filers to back up with ndmp, so i believe that is probably causing more problems ...
thanks, further hints welcome :)
05-19-2014 04:55 AM
I have also seen this...every weekend!
The filer/Netbackup have another attempt and the second iteration succeeds.
It only affects two vols - the same two - so you could say its data dependant.
Specifically its two folders within two vols ie /vol/vol1/folderY, /vol/vol2/folderZ.
Jim
05-19-2014 05:09 AM
hi jim!
can't confirm that it's always the same volumes - and i WISH that it would work the second attempt
but unfortunately it doesn't all the time, and i get problems with my backup-window when i have to backup 6TB via NDMP and need to wait for the second attempt ;)
what have you tried so far to get rid of the issue?