DR Test - new storage node, how to define device

Dave,

Here is another option that may help:

To scan tape drive through command line, change to the networker installation path and run the following command:

dvdetect -s

This information is from the EMC Knowledgebase article #esg119973: Manually discovering devices on server.

Mark

Mark,

Thanks, I have the 7.6 DR guide, here is where I’m failing for new node DRPFEE, I’ll look through the other document.

Info

Wednesday 10:18:48 AM

event

media

Media info: No configured libraries detected on storage node drpfee.

Info

Wednesday 10:18:48 AM

event

media

Media info: Could not find NSR device resource for rd=drpfee:DRPFEE$MKC400:

Info

Wednesday 10:18:48 AM

event

media

Media info: Found new LTO Ultrium-4 device DRPFEE$MKC400: on host . ()

Info

Wednesday 10:18:50 AM

event

media

Media info: restarting nsrmmd #1 on PROFEE now

Info

Wednesday 10:18:51 AM

event

media

Media info: restarting nsrmmd #1 on PROFEE in 2 minute(s)

Notice

Wednesday 10:18:51 AM

event

media

Media notice: check storage node: PROFEE (Host name verification failed - unknown host: 165.68.15.182)

Info

Wednesday 10:18:52 AM

event

media

Media info: Detection process for host drpfee reports dvdetect process failed on signal: Errno 99

Info

Wednesday 10:18:52 AM

event

media

Media info: Search for new backup devices completed with one failure.

Thanks,

Dave

Hello dkrakar!

I am very glad to see you are testing the DR process! Imagine what it would be like if your CIO was breathing down your neck because other systems needed to be restored!! I wish more people would take a proactive stand on DR testing!

I will admidt that I am not that familiar with openVMS.  However, the process should be basically the same regardless.

Please get the latest edition of the NetWorker Admin guide.  It can be located from powerlink at Support > Technical Documentation and Advisories > Software ~ J-O ~ Documentation > NetWorker Family > NetWorker > 7.6 & Service Packs (or the version you are running).

In this guide, look for the information for configuring a storage node. In that section you will find the instructions for adding a storage node to the environment.  There is also a section for Autodetection of libraries and devices that may help as well in getting the library and drives in place.

Let us know how things progress for you!

Mark

Mark,

Thanks, I’ve finally got the device running on OpenVMS, now I performed my scanner –s –v –m command and it say’s it rebuilt the index.

I’m trying to issue my recover command

$mc nsr$system:recover -s ucp-alaska -S 1237899742 -d $1$DKB2:[000000] –R drpfee -i "R"

Warning: could not establish job id with jobd. Continuing on.

not enough core

Could not connect to server.

If I do the same command from the Windows Server I get

C:\Program Files\Legato\nsr\bin>recover -s ucp-alaska -S 1237899742 -d $1$DKB2:[ 000000] -R drpfee -i "R"

Initiating remote recover to drpfee from server ucp-alaska, this may take a while...

Recovering files within DISK$SN_JRNDISK:[000000] into $1$DKB2:[000000] NSR server `ucp-alaska': no matching devices on `server'

recover: waiting 30 seconds then retrying ( I let this go for about an hour with this repeating)

If OpenVMS is my storage node and the Windows Server is the Networker Server and also a possible Storage Node, where should the tape drive be connected.

It’s an HP LTO-4 Ultrium 1730 (I think) with one drive, 7 slots and a mailslot.

Thanks,

Dave

Dave,

typically your storage node can be part of your Networker Server. I am almost sure that when you purchase your server licensing it includes the server as a storagenode.

Here is my stratedgy and if you like it feel free to adopt it.

My networker server is a Virtual Machine and i hav my Indexes and DB replicated Via SAN to my second Data Centre. (DR Covered)

I have 2 storage nodes. They are both seperate Physical boxes and live in each DC "Data Centre" I still hold the built in license for my server to be a Storage node but being a VM this does not support "Correctly" Attached SCSI devices. Now i don't want everyone jumping on this post and saying but ESX 4.1 does support direct attached SCSI. VMware say it does but i have had 23 sites with numourous problem with Direct attached tape drives so my advise is stay away from it.

If you have licensing and a need to keep your Library seperate or your environment is large or the Networker server is struggling with I/O Disk Cache reads and writes then seperate it.

Is your library fabric attached? If so you might need to do some zoning or add a HBA to allow connectivity.

Once you can see your SCSI device attahced to the Storage node "Persistant binding" Highly recommended. you should just have to rescan your SN in Networker and it should  pickup the Library.

I nnetworker you will most likelly need to still configure the library with Cleaning media slots and auto tape management. Also any other features you need or want. Idle device time out for an example.

Remember if your going to use a seperate storage nodeand plan to backup to this you will need to specify the storage node in your jobs and pools for access to the unit.

Best of luck hope this helps.

Scotty.

Mark,

I’m getting a lot closer, I put the Tape-drive back onto the Windows machine and re-installed Networker on my OpenVMS Server.

I can now recover files, but am getting “recover: no decryption key available!”.

If I go to my OpenVMS Server, the files are there in the correct directories, and have correct size and creation date stamp, but if I try to open/read a file, it acts like it is blank.

How do I get the “key” recovered or restored ?

DRPFEE (Disaster Recovery Server)

Total of 24 files, 24 blocks.

$ type MESSAGE1006717.TXT;2

$

PROFEE (Live Production Server)

PROFEE_$ type MESSAGE1006717.TXT;2

Interface # 304 REG PROCESSOR Crashed for group 3 in UCI: TEST43

PROFEE_$

Thanks,

Dave

Dkrakar,

I am sorry for the delay in my response. I have been travelling.

Let me look and see what needs to be done to recover the decryption key. I will post it as soon as I find it.

Mark

By the way, a couple of other notes:

When NetWorker  Server is installed, it also installs the Storage node and client peices.  This allows you to backup the NetWorker server without any other installs needing to be done.  It also is what allows the NW Server to manage the devices.

Also, as I always do, I STRONGLY recommend that you review the Disater Guide for the NetWorker version you have and have a plan in place for all evnironements in your facility (Exchange, SQL Oracle, Active Directory / LDAP).

Disaster Recovery plan is not just recovering the data, but knowing WHO needs to be involved, how to reach them and at what point in the process they need to be involved.

Upfront it takes time (and some arguing :-) ), but it i will save a LOT of time when it comes to actually recovering an environment!

The disaster recovery guide, along with several other manuals for version 7.6, can be found on PowerLink at Support > Technical Documentation and Advisories > Software ~ J-O ~ Documentation > NetWorker Family > NetWorker > 7.6 & Service Packs

Another guide that is well worth downloading and keeping handy is the command reference guide.  It has all of the NW commands that are used from cli and will give you examples, options, etc.

Mark

Mark,

Thanks, EMC Live Chat told me how to enter our encryption passphrase, now I just need to track down who initiated that and can remember what it is.

I have the DR Guide, the OpenVMS Admin Guide, but the command reference guide sounds good.

Thanks,

Dave

Well,  I have some bad news and some good news!

The good news (kind of)

This is a known issue.  It is addressed with a hotfix.

The bad news:

You need to call technical support to get the hotfix.

Refer them to EMC KB article esg118715

Let me know if you have any issues with that and I will take care of it.  Let me know what the service request number is if you do.

Thank you,

Mark

Okay, found the passphrase and entered that into the datazone field.  still getting error message, even when using -p option with "passphrase goes here" arguement.  Have case open with EMC and moved my arguements around in my recover command and still get same result.

recover -s -S -d -iR -R -p

recover -s -S -d -R -p -iR

recover -s -S -R -d -p -iR

Initiating remote recover to DRPFEE from server ucp-alaska, this may take a while...
  X = "OTHER"
Recovering files within DISK$SN_JRNDISK:[000000] into $1$DKB2:[000000]
$1$DKB2:[CACHE_JOURNAL]UOC20110430.002;1
$1$DKB2:[CACHE_JOURNAL]UOC20110430.002;1: file exists, renaming to $1$DKB2:[CACH
E_JOURNAL]UOC20110430.002_R;1
recover: no decryption key available!

So file is there, but I can't read, open, or type it.  It acts like it is blank.

Thanks,

Dave

I mentioned this and they couldn’t find the article.

My case number is 41015154, when I search I get something weird with that number.

Thanks,

Dave

Dave,

Did you mention the EMC KB article I mentioned above?

Refer them to EMC KB article esg118715

This is a hotfix that should resolve the issue.

Mark

Dave,

I have emailed the support rep with the information.  It is possible he needs to adjust his view on the support browser - it's an internal EMC thing.

Anyhow, while you are waiting for him, I would suggest that you try using the following command for the recover - this is the order the options should go in:

recover -S ssid -iR -R recover-target -c client -d destination -s server -p pass-phrase

I am NOT saying this will resolve the issue, but it might.

Let me know how you make out!

Mark

Dave,

No problem – let me know what happens!

Mark Bellows