Forcing VMWare Access Host

[Hollis_B]

This may be documented somewhere so let me apologize if it is, I may not have entered the right magical phrase into the google machine to get the right results.

My environment is made up of 3 Netbackup servers (7.5.0.6); 1 Linux master/media server, 1 Windows media server and 1 Linux media server. The only host with SAN access to the vm datastore is the Windows media server and is specified in the master server properties --> VMWare Access Host section. On the policy for the vm backups I have selected to use SAN and to use this windows media server but with all of these properties set it will still try to use the other 2 media servers if this one is busy. The resulting error when attempts are made from the wrong media server is status 59. Is there a way to keep this from happening?

I did find this note in the Symantec Netbackup for VMWare Administrators Guide - Page 52 but wasn't sure if it was my problem.

http://kbdownload.symantec.com/resources/sites/BUSINESS/content/live/DOCUMENTATION/5000/DOC5162/en_U...

=======================================================
Note:

The storage unit that is specified in the policy must be unique to

your Windows media servers. If the storage unit is also available on a UNIX

media server, the snapshot cannot succeed.

=======================================================
We are using SLPs that point to a single storage pool.

Thanks in advance.

[Mark_Solutions]

NetBackup should not do what you describe .. and if it did i would expect a 29 error and not a 59 error which is an access denied message.

It does sound like there is something wrong here

Could we see the outputs of the following fo rthis policy please:

bppllist policyname -U

bpplinfo policyname -U

Thanks

[Hollis_B]

Output provided in the attachment.  Thx

[Mark_Solutions]

Thanks - could i also see a screen shot of the VMware tab and also see a copy of the file \program files\veritas\netbackup\db\class\VM_WIN_PROD_1\info

Thanks

[Hollis_B]

CLIENT_TYPE 40
FOLLOW_NFS_MOUNTS 0
CROSS_MOUNT_POINTS 0
CLIENT_COMPRESS 0
PRIORITY 0
CLIENT_ENCRYPT 0
DISASTER_RECOVERY 0
MAX_JOBS_PER_CLASS 4
MAX_FRAG_SIZE 0
ACTIVE 0
EFFECTIVE_TIME 1304017478
COLLECT_TIR_INFO 0
EXTENDED_SECURITY_INFO 0
IND_FILE_RESTORE_FROM_RAW 1
BLOCK_INCREMENTAL 1
STREAMING 0
FROZEN_IMAGE 1
BACKUP_COPY 0
NUMBER_OF_COPIES 1
FAIL_ON_ERROR 0
CHECKPOINT 0
CHKPT_INTERVAL 0
ENABLE_PFI 0
OFFHOST_BACKUP 1
USE_ALT_CLIENT 0
USE_DATA_MOVER 0
DATA_MOVER_TYPE -1
COLLECT_BMR_INFO 0
RES_IS_SS 1
GRANULAR_RESTORE_INFO 0
JOB_SUBTYPE 0
USE_VIRTUAL_MACHINE 1
GENERATION 151
IGNORE_CLIENT_DIRECT 0
USE_ACCELERATOR 0
CLASS_ID 5D91AAF271CA11E08334C3B5AC14E3F7
ALT_CLIENT_NAME hp38nb02
SNAPSHOT_METHOD VMware_v2
SNAPSHOT_METHOD_ARGS file_system_optimization=1,snapact=2,drive_selection=0,Virtual_machine_backup=2,disable_quiesce=0,nameuse=1,exclude_swap=1,skipnodisk=0,post_events=1,trant
ype=san:nbd
EXCHANGE_SOURCE 3
APPLICATION_DISCOVERY 0
DISCOVERY_LIFETIME 28800000
ENABLE_META_INDEXING 0
RESIDENCE Daily_Inc_SLP_SSO_COLO *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
POOL NetBackup NetBackup NetBackup NetBackup NetBackup NetBackup NetBackup NetBackup NetBackup NetBackup
SHARE_GROUP *ANY*
DATA_CLASS *NULL*
#VMD5_DIGEST=bc 1e 25 6f 18 1b e3 40 6d bc af 72 56 58 96 55
 

[Mark_Solutions]

That all looks fine - so finally can we see the SLP and Storage Unit please:

Just wondering if the storage unit / pool if also available to other media servers so when it cal the storage unit it can call another media server .. if that happened it would use your VMware backup host to get the data over the SAN but it would be passed back to another media server to write it to disk / tape.

If the vmware backup host did not have the other media server in its list it would cause the 59 error - only thing i can think of that is causing this

[Hollis_B]

Sorry Mark but can you provide me the command for the SLP?

Label:                sso_nbuapp1_STU
Storage Unit Type:    Disk
Storage Unit Subtype: DiskPool (6)
Host Connection:      ssonbu
Host Connection:      hp38nb03
Host Connection:      hp38nb02
Concurrent Jobs:      65
On Demand Only:       yes
Max MPX:              1
Max Fragment Size:    51200 MB
Block Sharing:        yes
Ok On Root:           no
Disk Pool:            ssopool01
Snapshots:            no
Replication Primary:  no
Replication Source:   no
Replication Target:   no
Mirror            :   no
Independent       :   no
 

[Mark_Solutions]

The storage unit may be enough .. when it fails is it one of these two that does the job?:

Host Connection: ssonbu
Host Connection: hp38nb03
 

[Hollis_B]

Yes the failures occur on those two media servers.

Thx

[Mark_Solutions]

If you dont need them to have access to the storage unit them remove them but consider all implications ... but as per your original posting that will be your issue:

=======================================================
Note:

The storage unit that is specified in the policy must be unique to

your Windows media servers. If the storage unit is also available on a UNIX

media server, the snapshot cannot succeed.

=======================================================

[Hollis_B · Accepted solution]

Thanks Mark for your time, I was hoping that was not my problem.