Netapp filers running 8.2 cluster mode

[bkup1234]

Hello All,

I need some information regarding  NDMP backups . I ahve a netapp filer running on 8.2 cluster mode  which we want to backup using remote ndmp.As per the HCL guide it is supported for versions 7.x. Our master was recently upgraded to 7.5.0.6. At present we are trying to backup the cluster as individual nodes currently under test. The credentials were added successfull for one of the filer nodes. And a test backup policy was setup , but the backup fails with the following error , any idea what is goingon here.

1/15/2014 13:53:12 - positioned DK0286; position time: 0:00:01
01/15/2014 13:53:12 - begin writing
01/15/2014 13:54:28 - Error ndmpagent (pid=21412) ndmp_data_connect_v3 failed, status = -1 (-1)
01/15/2014 13:54:28 - Error ndmpagent (pid=21412) NDMP backup failed, path = 192.168.200.1:/netappcd1/test
01/15/2014 13:54:28 - Info bptm (pid=21446) EXITING with status 99 <----------
01/15/2014 13:54:28 - Info ndmpagent (pid=0) done. status: 99: NDMP backup failure
01/15/2014 13:54:28 - end writing; write time: 0:01:16
NDMP backup failure  (99)

Connecting to host "192.168.200.1" as user "root"...
Waiting for connect notification message...
Opening session--attempting with NDMP protocol version 4...
Opening session--successful with NDMP protocol version 4
  host supports TEXT authentication
  host supports MD5 authentication
Getting MD5 challenge from host...
Logging in using MD5 method...
Host info is:
  host name "bminetappc-01"
  os type "NetApp"
  os version "NetApp Release 8.2P5 Cluster-Mode"
  host id "1874011246"
Login was successful
Host supports LOCAL backup/restore
Host supports 3-way backup/restore
 

[Will_Restore]

Pathname should be something like

/vol/dir not ip:/mountpoint

[bkup1234]

Tried /vol/bminetappcd1/test , the backup continues to fail with ec-99.

01/15/2014 20:55:14 - Error ndmpagent (pid=29071) ndmp_data_connect_v3 failed, status = -1 (-1)
01/15/2014 20:55:14 - Error ndmpagent (pid=29071) NDMP backup failed, path = /vol/bminetappcd1/test
01/15/2014 20:55:14 - Info bptm (pid=29113) EXITING with status 99 <----------
01/15/2014 20:55:14 - Info ndmpagent (pid=0) done. status: 99: NDMP backup failure
01/15/2014 20:55:14 - end writing; write time: 0:01:16
NDMP backup failure  (99)

[Will_Restore]

Check on the filer itself for valid path and use that in NetBackup NDMP policy

[Will_Restore]

NetBackup Compatibility Guide says

Data ONTAP version 8.x (7-mode) supports MultiStore (vFiler) with
Remote NDMP and 3-way NDMP connection methods. --- Data
ONTAP version 8.x (7-mode) supports NetBackup Advanced Client
NAS Snapshot. --- Supports wildcard characters for backup selections
lower than the volume level. --- The NetApp Cluster Aware Backup
(CAB) NDMP extension introduced in Data ONTAP 8.2 is not
supported. Tape devices attached to the NetApp are only supported in
node-scope-mode. A backup of a vServer as an NDMP client will be a
3-way backup.

http://kbdownload.symantec.com/resources/sites/BUSINESS/content/live/TECHNICAL_SOLUTION/76000/TECH76495/en_US/nbu_7x_hcl.pdf?__gda__=1389892083_b7b5e18a31d2d4730907f81bfdcfe95c

[bkup1234]

Wr,

I am  trying to backup the filers as a node not as a cluster , just like a non-cluster filer. The backup still fails with the error 99. Any thoughts?

[rk1074]

please make sure port 10000 is open bidirectionally on the Filer..NDMP requires it to be working..

[rk1074]

also,

Since you said its a Remote NDMP,o n the media server, test the connection to the NAS device on the NDMP port (10000) by typing the telnet command. Try both the hostname and its IP address. For example: telnet ndmp_host 10000

[mnolan]

Hello,

Usually what I see for the error,

"ndmp_data_connect_v3 failed",

is that the filer cannot reach back to the media server for the secondary connection to pass data.

Per the NetBackup 6.x and 7.x firewall port requirements technote at http://www.symantec.com/docs/TECH136090 ,

"The SERVER_PORT_WINDOW is used inbound from the filer to the media server for remote NDMP and can also be used for efficient catalog file (TIR data) movement with local and 3-way NDMP. "

By default the server port windows is the range of 1024-5000 and a random port will be opened in this range that the filer will be told to connect to. Therefore, in addition to port 10000 to be open between the filer and media server, we also require the server port window range to be open.

[jim_dalton]

And for good measure 'ls' the vol in question. (The first err is the comms err, so the vol probably is ok but ls would prove it).

Its an advanced cmnd in ontap (for some reason).

Jim

[Pagarodog]

If you take a look at the HCL for Netbackup dated 2014-01-31 page 198, in the NetApp section, you will see >>

"The NetApp Cluster Aware Backup (CAB) NDMP extension introduced in Data ONTAP 8.2 is not supported.  Tape devices attached to he NetApp are only supported in node-scope-mode.  A backup of a vServer as an NDMP client will be a 3-way backup."

All is not lost.  Turn on Node-Scope-Mode (NSM). Turn off clear text password and Turn on encrypted for NSM.

You need to setup DNS and reverse DNS for the nodes and connect to those addresses as the NDMP hosts and only ask NetBackup to only backup the shares that the node hosts.  You will be connecting to the management LIF for the node.

Setup Credentials>NDMP HOSTS>for each node

Check that you are connecting to the nodes:

MASTERSERVER:/usr/openv/volmgr/bin # ./tpautoconf -verify node-store1

Connecting to host "node-store1" as user "root"...

Waiting for connect notification message...

Opening session--attempting with NDMP protocol version 4...

Opening session--successful with NDMP protocol version 4

  host supports MD5 authentication

Getting MD5 challenge from host...

Logging in using MD5 method...

Host info is:

  host name "node-store1"

  os type "NetApp"

  os version "NetApp Release 8.2.1RC1 Cluster-Mode"

  host id "5555555555"

Login was successful
Host supports LOCAL backup/restore
Host supports 3-way backup/restore

You can also issue the-- MASTERSERVER:/usr/openv/volmgr/bin # ./tpautoconf -probe node-store1 to check for the existence of tape drives although this gives every possible incarnation of such devices.

Configure your storage devices making sure to select NDMP for the host/nodes you setup above.

Setup an NDMP policy to backup the nodename (host nodestore1 for example) using a path you created when you configured the new storage paths.

Seems to obey NDMP rules for setting the EXCLUDE variable.  This is how I fragment all backups that exceed 5 TB.

If you connect to the "cluster", you will get results like you have shown.  You must connect to the nodes (for now).

Good Luck.

P

[nomecks]

I don't know if you figured this issue out yet, but it looks like your backup path is wrong. With node level backups, your client should be the cluster node the vserver is living on, and the backup path should be vserver/volume. If you don't know which cluster node the data lives on, just add them all, then run a test backup. All the backups except for the correct node will fail. So if you have clusternode-01 and clusternode-02, vserver "backupvserver" and a volume called "volume", then your clients would be clusternode01 and 02, and the volume path would be /backupverserver/volume. /vol/ is not generally used in cluster mode, unless you specifically junction volumes into /vol.