This post is more than 5 years old
6 Posts
0
3956
Networker Upgraded to 8.1.1 NetApp NDMP dump proxy save no longer works
I have upgraded our networker from 7.4.5 to 8.1.1.2 this past few weeks, with the inclusion of upgrading the OS from win2003 to win2008. While it was on win2008 and 7.4.5, the NDMP via proxy worked flawlessly, now it ignores "i think" -M -P "other_ipaddress" setting.
So the server has two nics, one "regular_ip_address" and one "other_ipaddress" for ndmp dumps.
The command that used to work is:
nsrndmp_save -M -T dump -P "other_ipaddress"
Now on the new version this same command does not respect the "other_ipaddress" everything goes through the "regular_ipaddress" which creates a huge bottleneck.
The server is a standalone networker, no other storage nodes. So the proxy was pointing to the other interface.
Any suggestions? Is this a bug in 8.1.1.2?
The NetApp is on dataontap 8.something.
Thanks
rolajos1
6 Posts
1
June 8th, 2014 21:00
With the help of EMC tech, it should have worked with the ipaddress.
With him working on it for an hour he figured out the solution and that is:
nsrndmp_save -M -T dump -P
And in the Application Information of the client add the
NSR_DSA_SAVE_SERVER=
And this is still on 8.1.1.2
Thanks for all the suggestions.
ble1
2 Intern
2 Intern
•
14.3K Posts
0
March 10th, 2014 17:00
I use 8.0.3.1 and it works there. I would argue that this is a bug in case that -P parameter is ignored.
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
0
March 11th, 2014 01:00
Hola, try using NSR_DSA_NODE=
https://support.emc.com/search?text=NSR_DSA_NODE
rolajos1
6 Posts
0
March 11th, 2014 10:00
No change
with the NSR_DSA_NODE
I have even changed the hosts files on both netapp and the backup server and only, from what I can tell, the small amount of traffic that has gone through the "other_ipaddress" might be the just metadata, about 1GB or so, after that I can clearly see that it is going through the "regular_ipaddress" again.
So the hostname has little to say, if the network is still reachable. But unfortunately I can't VLAN separate this since this is a production filer.
rolajos1
6 Posts
0
March 11th, 2014 12:00
Just an FYI
5183:nsrndmp_save: DSA is listening for an NDMP data connection on: "regular_IPAddress", port = 8839
42952:nsrndmp_save: xxxxxxxx:/vol/vol_xxxxxxx/ NDMP save running on 'xxxbackupserverhostnamexxxxx'
86724:nsrdsa_save: DSA listening at: host 'other_ipadress', IP address 'regular_ipaddress', port '8839'.
42958:nsrdsa_save: Performing Immediate save
I will also try to update the server tomorrow to 8.1.1.3
rolajos1
6 Posts
0
March 11th, 2014 13:00
It was defined a long time ago through the front-end
ble1
2 Intern
2 Intern
•
14.3K Posts
0
March 11th, 2014 13:00
Here is how it is in my case using 8.0.3.1 (Linux) - also with NetApp.
How did you define filer - over front-end or back-end address?
ble1
2 Intern
2 Intern
•
14.3K Posts
0
March 11th, 2014 14:00
That would be the only difference in setup I guess when comparing to what I have. You can always rename client to see if that makes any changes if update won't help (I suspect it won't as I didn't see anything described as you have on fix list).
Thierry101
2 Intern
2 Intern
•
326 Posts
0
March 19th, 2014 00:00
WHat is entry in storage node field..was informed that -P is irrelevant and storage node field supposed to take care of that but was told in 7.6x it doesnt work and the workaround is to put the -P in command....but in 8 was told that the sn field shd work without p...no?
rolajos1
6 Posts
0
March 20th, 2014 10:00
i had nsrserverhost, which i have changed to the "other_Ipaddress", but no change in behavior.
Still on 8.1.1.2 since i could not get the upgrade started with the 8.1.1.3 installer.
ble1
2 Intern
2 Intern
•
14.3K Posts
0
March 20th, 2014 16:00
8.1.1.2 -> 8.1.1.3 is not upgrade, but update and as for such you will to uninstall 8.1.1.2 first and then install 8.1.1.3. I personally doubt 8.1.1.3 will fix your issue as I didn't see anything in fix list specific to this problem.
sabogal_juan
1 Message
0
May 18th, 2016 10:00
Thanks rolajo,
The answer was great help for me, The networker server has two Vlans, a is administrative and another is of backup.
Always the backup used the IP of vlan administrative and the backup failed
administrative IP = 10.181.0.15
##############################################################
82966:nsrndmp_save: Using networker as the NetWorker server.
42910:nsrndmp_save: Performing Non-DAR Backup..
83564:nsrndmp_save: Performing full backup
42794:nsrndmp_save: Performing backup to Non-NDMP type of device
42658:nsrdsa_save: DSA savetime = 1463584832
85183:nsrndmp_save: DSA is listening for an NDMP data connection on: 10.181.0.15, port = 9778
42597:nsrndmp_save: data connect: failed to establish connection
42886:nsrndmp_save: Unable to start the NDMP backup process.
42617:nsrndmp_save: NDMP Service Log: SELECTED NODE: 172.30.80.154 WWN 2009000b0804a6fa for BE session
42619:nsrndmp_save: NDMP Service Error: Cannot connect to the mover
####################################
So the solution was:
backup command:
nsrndmp_save -M -T dump -P
appl:
NSR_DSA_SAVE_SERVER=