cancel
Showing results for 
Search instead for 
Did you mean: 

VMware on SAN backup

T_N
Level 6

Hi All;

Anyone know how to set up VMware on SAN Backup. The outstanding question was why the Media Servers were not handling the SAN-based backups correctly and instead defaulting to IP-based backups.  Thank you.

Tom

1 ACCEPTED SOLUTION

Accepted Solutions

vtas_chas
Level 6
Employee

Not to be flippant, but if you can't see the LUNs, it isn't setup correctly.

This isn't just a zoning issue, it's a masking issue, too.  Masking tends to trip people up the most.

Tell the SAN guys to configure the zoning for the appliance or whatever your bakup host is EXACTLY like they would another ESX host.

Charles
VCS, NBU & Appliances

View solution in original post

15 REPLIES 15

vtas_chas
Level 6
Employee

There are a number of reasons that this can happen.  The most common one I've seen has been incorrect zoning and LUN Masking of the media server or appliance to the datastores.  Like 9 out of 10 common.

Can you give us some details about your configuration and what you've done to troubleshoot?

Charles
VCS, NBU & Appliances

Thank you for your reply. VMware guy said all the blade chassis are properly connected.


Based upon the logs, it appears to be a zoning issue.   Information from the logs is included below:

 

   06/27/2017 15:39:28 : g_vixInterfaceLogger:libvix.cpp:1797 <DEBUG> : [VFM_ESINFO] 2017-06-27T15:39:28.487-04:00 error -[7F5F3DB9F720] [Originator@6876 sub=Default] No path to device LVID:5891e320-eacee65c-ad03-0025b50a006e/5891e320-cf8d96e4-832f-0025b50a006e/1 found.
    06/27/2017 15:39:28 : g_vixInterfaceLogger:libvix.cpp:1797 <DEBUG> : [VFM_ESINFO] 2017-06-27T15:39:28.487-04:00 error -[7F5F3DB9F720] [Originator@6876 sub=Default] Failed to open new LUN LVID:5891e320-eacee65c-ad03-0025b50a006e/5891e320-cf8d96e4-832f-0025b50a006e/1.

 


What command lines to check if they set up correctly zoning and LUN Masking of the media server or appliance to the datastores

Thiago_Ribeiro
Moderator
Moderator
Partner    VIP    Accredited

Hi,

You need to guarantee that vmware backup host can see the Luns.To check from Appliances, attached there is an example.

Here it is a good document to help you

VMware Transport Modes: Best practices and troubleshooting - https://www.veritas.com/support/en_US/article.TECH183072

 

Regards,

 

Thiago

Thank you for your documentation. I have couple questions:

1)      How to present SAN Volumes to netbackup Appliance ? Because I read on the article “ NetBackup 7 Configuration for VMware vStorage Backups “ they said “Present SAN Volumes to the NetBackup Windows Media Server (Backup Host)”

Before the Zone with the Data Store is made make sure that automount is disabled and Scrubed on your media server. How can I disable automount & scrube on netbackup appliance ?

https://  www.virtualdennis.com/netbackup-7-configuration-for-vmware-vstorage-backups

2) SAN & VMware guys they said they already zong between media server (netbackup appliance) and data Store.  What command lines to check if they set up correctly zoning and LUN Masking of the netbackup appliance (media server) to the datastores?

 

Thank you.

vtas_chas
Level 6
Employee

Not to be flippant, but if you can't see the LUNs, it isn't setup correctly.

This isn't just a zoning issue, it's a masking issue, too.  Masking tends to trip people up the most.

Tell the SAN guys to configure the zoning for the appliance or whatever your bakup host is EXACTLY like they would another ESX host.

Charles
VCS, NBU & Appliances

Thiago_Ribeiro
Moderator
Moderator
Partner    VIP    Accredited

Hi @T_N

Take a look


1) How to present SAN Volumes to netbackup Appliance ? Because I read on the article “ NetBackup 7 Configuration for VMware vStorage Backups “ they said “Present SAN Volumes to the NetBackup Windows Media Server (Backup Host)”

A: You must present SAN Volumes to Netbackup Appliance. I use SAN Transport Mode, in my case I sent to SAN guy the WWNs address of Appliance HBAs, look

**** Ports ****
Bus ID Slot Port Port WWN Status Physical State Configuration Type Speed Remote Ports
06:00.0 Slot5 1 21:00:00:24:FF:7C:ED:AA Online Initiator Target(FTMS) 8 gbit/s 0x500601680864586c
0x500601600864586c
06:00.1 Slot5 2 21:00:00:24:FF:7C:ED:AB Online Initiator Initiator 8 gbit/s 0x2005000e1116afb7
07:00.0 Slot6 1 21:00:00:24:FF:7C:EE:78 Online Initiator Target(FTMS) 8 gbit/s 0x5006016a0864586c
0x500601620864586c
07:00.1 Slot6 2 21:00:00:24:FF:7C:EE:79 Online Initiator Initiator 8 gbit/s 0x2005000e1116afb7
81:00.0 Slot4 1 21:00:00:24:FF:7C:EF:10 Online Initiator Initiator 8 gbit/s 0x2002000e1116afb7
81:00.1 Slot4 2 21:00:00:24:FF:7C:EF:11 Online Initiator Target(MSDP) 8 gbit/s 0x500601610864586c
0x500601690864586c
82:00.0 Slot2 1 21:00:00:24:FF:7C:F2:28 Online Initiator Initiator 8 gbit/s 0x2002000e1116afb7
82:00.1 Slot2 2 21:00:00:24:FF:7C:F2:29 Online Initiator Target(MSDP) 8 gbit/s 0x5006016b0864586c
0x500601630864586c

Before the zooning is done the field Remote Port was blank...After done that information appear, and after run the "Show" commnad, I saw the Luns


Before the Zone with the Data Store is made make sure that automount is disabled and Scrubed on your media server. How can I disable automount & scrube on netbackup appliance ? - https:// www.virtualdennis.com/netbackup-7-configuration-for-vmware-vstorage-backups

A: From what i understand reading your article Its is applied for Windows 2008 R2 hosts, and sincerily I dont know how to do that..sorry!

2) SAN & VMware guys they said they already zong between media server (netbackup appliance) and data Store. What command lines to check if they set up correctly zoning and LUN Masking of the netbackup appliance (media server) to the datastores?


A: If you already sent the WWNs address and you still can not see the Luns, try to reboot the Appliance...I do this, and it worked for me after reboot.

Take a look this TN - https://www.veritas.com/support/en_US/article.000100581

 

Regards,

 

Thiago

I sent the FC output from netbackup appliance to Storage team.After they check,  it's working now. But when I run 2 vmware backup jobs (2 vmware clients) using to different netbackup appliance (dd5 & dd6), their speeds are difference 13906 kb/sec and 35993 kb/sec. Any idea why ? Thank you.

 

Thiago_Ribeiro
Moderator
Moderator
Partner    VIP    Accredited
Hi,
The zooning was make for both appliances? Can you confirm? Other thing, you can check the transport type from detailed status from Job. Tale a look.

https://www.veritas.com/support/en_US/article.000018735

Regards,

Thiago

Yes they did on both appliances, I am able to see backup jobs work well but only speed, the speeds are differents on both appliances

 

07/24/2017 07:33:35 - Info bptm (pid=34148) start backup
07/24/2017 07:33:41 - Info bpbkar (pid=34147) INF - Transport Type =  san
07/24/2017 07:43:37 - Info bpbkar (pid=34147) bpbkar waited 34 times for empty buffer, delayed 8094 times

vtas_chas
Level 6
Employee

Are you using resource limits?  Normally wide variation in performance can be helped by using the Resource Limits feature.

Charles
VCS, NBU & Appliances

I am not following your question "Are you using resource limits? ". I am using SLP that points to media server (netbackup appliance). Do we have any log to check performance on netbackup appliance ? Thank you.

Tom

vtas_chas
Level 6
Employee

Have you reviewed the Admin Guide for VMware?  Looking at logs at this point is not likely to help you much if you're not following the best practices for configuring VMware VADP backups first.  There's no log that is going to say 'this is why you're not getting the same performance on two different appliances', for sure :-).  To set resource limits for VMware or Hyper-V, you look at the Host Properties -> Master Servers -> your master and look at Resource Limits in the list of properties.

You could have a whole host of things impacting this, from storage, to SAN, to ESXi host, to other backup jobs, etc.  I've seen folks that jump into VADP SAN transport backups saturate a datastore or ESXi host by not using resource limits in the policy.  Are you using the automatic selection queries to drive the policies or manual selections?  I've also seen systems that are already at the tipping point get a bunch of VMware SAN Transport backups thrown at them (again without resource limits) that sends them over the top and the appliance starts having issues.

There's lots of things here we don't know so I can't just point you to a single log to look at.

I suppose I should also get clarification on what version you're using.  It looks like you tagged your post 7.7.  Am I correct in assuming that's your version?  

Charles
VCS, NBU & Appliances

Thiago_Ribeiro
Moderator
Moderator
Partner    VIP    Accredited

Hi @T_N,

Can you see this TN? Maybe it can help you

How to use NBPerfchk to troubleshoot performance issues related to disk io and network transfer rates - https://www.veritas.com/support/en_US/article.000095782

 

Regards,

 

Thiago

Marianne
Level 6
Partner    VIP    Accredited Certified