Michael Leone
Advice: upgrading from 8.x to latest (18.x?)
May 06, 2019 08:59AM
OK, I absolutely have to get this upgrade out of the way, so I can contact
support if I need to. So what I am looking for is some advice, on how best
to proceed, in my situation.

Currently:
Physical Win 2008 R2 NW 8.2.3.4 server, with 1 LTO-6 tape library (8
drives) , attached by Fiber Channel. Seeing as to how the library is FC
connected, and can't run iSCSI (I checked), this means I am stuck with a
physical machine that can talk to those drives.
Backup jobs write to a Data Domain device, and then scheduled clone jobs
write the data to the tape drives. About 175 clients or so.

Mind you, everything else already is, or hopefully soon will be,
virtualized.

Here's where I need the advice: How do you proceed, in a situation like
this? Is it possible to make the new NW server a VM, and just leave the
current physical box as a storage node, and change all the backup jobs to
send their data to the storage node, as opposed to the NW server directly?
That should allow me to continue to use this type of library, I think.

Anybody have a link to a good blog site/YouTube video on the migration
process to whatever the latest version is (18.x, right? They just skipped
from 9 to 18?)?

I'm vaguely recalling the old round about way of migrating to a new host.
What I REALLY, REALLY DON'T want to do is the full recover of the mm and
CFI. I've got literally hundreds of gigs worth of entries in my CFI index
folder, and waiting to import them via a nsrck -L7 command doesn't bear
thinking about.

I recall being able to install a new version, rename the new empty mm, res
and index folders; copy in my old ones; restart.

But you can't do that with a differently named server, can you? Or do you
shut down the old services; rename the old server; roll out a new VM with
the old name and IP; and do the copy method I listed above?

Thanks for any insights. I've done a lot fo upgrades over the years, but
not migration to a new host (i vaguely recall we did that 10 years ago,
but I don't remember how, just now). Nor did I have this new wrinkle,
about the tape library being FC attached, but the NW server may not be.


--
Michael Leone
Network Administrator, ISM
Philadelphia Housing Authority
2013 Ridge Ave
Phila, PA 19121-4113
Hours: 8AM - 4PM EST
Tel: 215-684-4180
Cell: 215-252-0143
<mailto:michael.leone@pha.phila.gov>




--
This list is hosted as a public service at Temple University by Stan Horwitz
If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu
This message was imported via the External PhorumMail Module
Mark Leese
Re: Advice: upgrading from 8.x to latest (18.x?)
May 06, 2019 08:59AM
Hi Michael,

You will probably have to go through the NetWorker server rename procedure
as that would be the safest thing to do.

There are ways and means but they do carry a risk and would probably leave
you outside of NetWorker support.

Personally I would always ensure the nsr folder is on a SAN or dual
attached array so the volume can be moved between servers easily. I would
then follow the server rename procedure in the manual or the KB articles
from the support site.

Hope this helps!

Mark.

On Mon, 6 May 2019, 16:45 Michael Leone, <Michael.Leone@pha.phila.gov>
wrote:

> OK, I absolutely have to get this upgrade out of the way, so I can contact
> support if I need to. So what I am looking for is some advice, on how best
> to proceed, in my situation.
>
> Currently:
> Physical Win 2008 R2 NW 8.2.3.4 server, with 1 LTO-6 tape library (8
> drives) , attached by Fiber Channel. Seeing as to how the library is FC
> connected, and can't run iSCSI (I checked), this means I am stuck with a
> physical machine that can talk to those drives.
> Backup jobs write to a Data Domain device, and then scheduled clone jobs
> write the data to the tape drives. About 175 clients or so.
>
> Mind you, everything else already is, or hopefully soon will be,
> virtualized.
>
> Here's where I need the advice: How do you proceed, in a situation like
> this? Is it possible to make the new NW server a VM, and just leave the
> current physical box as a storage node, and change all the backup jobs to
> send their data to the storage node, as opposed to the NW server directly?
> That should allow me to continue to use this type of library, I think.
>
> Anybody have a link to a good blog site/YouTube video on the migration
> process to whatever the latest version is (18.x, right? They just skipped
> from 9 to 18?)?
>
> I'm vaguely recalling the old round about way of migrating to a new host.
> What I REALLY, REALLY DON'T want to do is the full recover of the mm and
> CFI. I've got literally hundreds of gigs worth of entries in my CFI index
> folder, and waiting to import them via a nsrck -L7 command doesn't bear
> thinking about.
>
> I recall being able to install a new version, rename the new empty mm, res
> and index folders; copy in my old ones; restart.
>
> But you can't do that with a differently named server, can you? Or do you
> shut down the old services; rename the old server; roll out a new VM with
> the old name and IP; and do the copy method I listed above?
>
> Thanks for any insights. I've done a lot fo upgrades over the years, but
> not migration to a new host (i vaguely recall we did that 10 years ago, but
> I don't remember how, just now). Nor did I have this new wrinkle, about the
> tape library being FC attached, but the NW server may not be.
>
>
> --
> Michael Leone
> Network Administrator, ISM
> Philadelphia Housing Authority
> 2013 Ridge Ave
> Phila, PA 19121-4113
> Hours: 8AM - 4PM EST
> Tel: 215-684-4180
> Cell: 215-252-0143
> <mailto:michael.leone@pha.phila.gov <michael.leone@pha.phila.gov>>
>
>
> -- This list is hosted as a public service at Temple University by Stan
> Horwitz If you wish to sign off this list or adjust your subscription
> settings, please do so via
> http://listserv.temple.edu/archives/emc-dataprotection-l.html If you have
> any questions regarding management of this list, please send email to
> owner-emc-dataprotection-l@listserv.temple.edu
>


--
This list is hosted as a public service at Temple University by Stan Horwitz
If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu
This message was imported via the External PhorumMail Module
Re: Advice: upgrading from 8.x to latest (18.x?)
May 06, 2019 01:50PM
We just did a very similar upgrade/migration just a few months ago with 2 major differences:
- we already had 2 storage nodes in place
- we have already migrated the tape save sets to our Data Domain cloud tier

This is the coarse sequence I would use:
- update your NW server from 8.x to 9.x first - I would never directly go from 8.x to 18.x.
- familarize with the new configuration and make sure that the system will run stable before you proceed
Do not forget licensing!
Allow some weeks to do this thoroughly enough
- define your new VM - same hostname name and ip
- shutdown NW and disable the NW client service
- use Robocopy to copy the disks to vmx files which you can later mount to the VM
- shutdown the old NW server
- add the disks to the new server configuration
- start the new NW server and activate the network
- make sure that it has the same SID
- re-enable the NW services and start NW
- make sure the system comes up nicely (but it will not totally correct, according to the wrong SN list)

On the old NW server ...
- de-install NW
- rename the host
- Install NW SN software

On the new NW server ...
- add the new SN
- re-create/edit you jukebox/device configuration
- adjust the SN list for the clients
- Make extensive tests and monitor your system closely
- Later: upgrade the NW server software to 18.x

The downside is that you do not have the old NW server any more in case something goes wrong.
That's why I personally would establish an additional SN first so that you have your old NW server in case the procedure will fail.
Michael Leone
Re: Advice: upgrading from 8.x to latest (18.x?)
May 08, 2019 06:59AM
> Hi Michael,
>
> You will probably have to go through the NetWorker server rename
> procedure as that would be the safest thing to do.
>
> There are ways and means but they do carry a risk and would probably
> leave you outside of NetWorker support.
>
> Personally I would always ensure the nsr folder is on a SAN or dual
> attached array so the volume can be moved between servers easily. I
> would then follow the server rename procedure in the manual or the
> KB articles from the support site.


Yeah, I was afraid of that.

What of the others points of consideration - can you have/is it
recommended to have the NW server itself be a VM? In my case, since I need
to keep writing to tape, and my tape library won't work via iSCSI, I need
to keep it connected via FC. Which means (I think) I need to make a
storage node that is FC connected, and define it to use the library. And
therefore change all my clone jobs to write to that library.

What are the rest of you folks doing? Are you still keeping the NW server
physical, and attached to tape storage of whatever sort? Are you using
virtual backups (meaning that you aren't backing up the "traditional" way,
an agent installed on each VM, etc)?

Thanks for all your help and advice, everyone!


--
This list is hosted as a public service at Temple University by Stan Horwitz
If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu
This message was imported via the External PhorumMail Module
Re: Advice: upgrading from 8.x to latest (18.x?)
May 08, 2019 08:49AM
As we had to replace our Centera by an ECS anyway, we decided to also migrate our 16-year old GRAU silo tape library to a disk based solution (private DD/ECS cloud).

So we migrated about 1000 LTO5 tapes. Fortunately I could use 12 drives in parallel to copy our clone these save sets. This went smoothly but it took a while. Especially if you have multiplexed tapes. We had luck because we only had about 25 of such tapes.

Then we migrated the backup server from a physical to a virtual machine. Since then we might get rid of our storage nodes which we will potentially scrap in the future.

About 95% of our servers (90% are VMs) are still beeing backed up the old fashioned way via a NW client (and a db module, where necessary). We do not use NMM for SQL backups as we habe good SQL admins who export the data to backup files.
Besides that we have not even setup 'fancy' clients using PSS and CBT backups (except for Exchange which does not work any other way).

We only have a few appliances that we backup via NVP.

BTW - we are using scripted cloning because it cuts the runtime of a workflow/group. Especially if you use tape cloning. However, now as a clone from the same DD Active Tier to DD Cloud Tier only transfers metadata, it just takes a few minutes every day instead of almost a whole day.
Mark Davis
Re: Advice: upgrading from 8.x to latest (18.x?)
May 08, 2019 08:59AM
Yeah, I was afraid of that.

What of the others points of consideration - can you have/is it recommended to have the NW server itself be a VM? In my case, since I need to keep writing to tape, and my tape library won't work via iSCSI, I need to keep it connected via FC. Which means (I think) I need to make a storage node that is FC connected, and define it to use the library. And therefore change all my clone jobs to write to that library.

What are the rest of you folks doing? Are you still keeping the NW server physical, and attached to tape storage of whatever sort? Are you using virtual backups (meaning that you aren't backing up the "traditional" way, an agent installed on each VM, etc)?

Thanks for all your help and advice, everyone!
========================================================

Michael,

Our Networker Server and Storage Node are both physical. We have a dual fabric FC SAN for our AFTD's. Our tape libraries, Quantum i500's, are also FC connected. Lots of redundancy in these connections. Plus FC just plain works better IMHO. We keep primary copies of backups on AFTD, with clones going to tape in our second Data Center. Data on disk in one Data Center is always cloned to the tape library in the other data center. This is true for both the Server and the Storage Node.

For our VMware world, we use Veeam. It's a great product for that area. I would be concerned if we were backing up VM's with a Networker server that was also a VM. If you lose then entire VM environment, where do you start?

Thanks my $0.02

Mark
Western University
London, ON




--
This list is hosted as a public service at Temple University by Stan Horwitz
If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu
This message was imported via the External PhorumMail Module
Re: Advice: upgrading from 8.x to latest (18.x?)
May 08, 2019 09:04AM
Mark made a very good comment on whether a NW server should be virtualized.
I would also support his thoughts if you think about an absolute desaster.

However, going virtual was our management decision.
Re: Advice: upgrading from 8.x to latest (18.x?)
May 08, 2019 11:59AM
In regard to: Re: [EMC-DataProtection-L] Advice: upgrading from 8.x to...:

> What of the others points of consideration - can you have/is it
> recommended to have the NW server itself be a VM? In my case, since I need
> to keep writing to tape, and my tape library won't work via iSCSI, I need
> to keep it connected via FC. Which means (I think) I need to make a
> storage node that is FC connected, and define it to use the library. And
> therefore change all my clone jobs to write to that library.

We're not doing that here, but I'm pretty sure that a physical storage
node + a VM NetWorker server is a supported configuration. As long as
you've designed your VM environment with the necessary I/O bandwidth
for indexes and the other I/O needs of a server, you should be fine.

Tim
--
Tim Mooney Tim.Mooney@ndsu.edu
Enterprise Computing & Infrastructure 701-231-1076 (Voice)
Room 242-J6, Quentin Burdick Building 701-231-8541 (Fax)
North Dakota State University, Fargo, ND 58105-5164


--
This list is hosted as a public service at Temple University by Stan Horwitz
If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu
This message was imported via the External PhorumMail Module
Preston de Guise
Re: Advice: upgrading from 8.x to latest (18.x?)
May 08, 2019 01:59PM
The NetWorker server can be virtualised; however, bear in mind it still needs to have a backup device to write to for bootstrap backups: in the case of a virtualised NetWorker server this should be an AFTD or Data Domain Boost device. Passthrough of tape to virtual machines isn't ever recommended. Even VMware themselves are pretty leery of it.

Cheers,
Preston.

Sent from my iPad

> On 9 May 2019, at 04:42, Tim Mooney <Tim.Mooney@ndsu.edu> wrote:
>
> In regard to: Re: [EMC-DataProtection-L] Advice: upgrading from 8.x to...:
>
>> What of the others points of consideration - can you have/is it
>> recommended to have the NW server itself be a VM? In my case, since I need
>> to keep writing to tape, and my tape library won't work via iSCSI, I need
>> to keep it connected via FC. Which means (I think) I need to make a
>> storage node that is FC connected, and define it to use the library. And
>> therefore change all my clone jobs to write to that library.
>
> We're not doing that here, but I'm pretty sure that a physical storage
> node + a VM NetWorker server is a supported configuration. As long as
> you've designed your VM environment with the necessary I/O bandwidth
> for indexes and the other I/O needs of a server, you should be fine.
>
> Tim
> --
> Tim Mooney Tim.Mooney@ndsu.edu
> Enterprise Computing & Infrastructure 701-231-1076 (Voice)
> Room 242-J6, Quentin Burdick Building 701-231-8541 (Fax)
> North Dakota State University, Fargo, ND 58105-5164
>
>
> --
> This list is hosted as a public service at Temple University by Stan Horwitz
> If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
> If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu


--
This list is hosted as a public service at Temple University by Stan Horwitz
If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu
This message was imported via the External PhorumMail Module
Re: Advice: upgrading from 8.x to latest (18.x?)
May 15, 2019 01:59PM
I have seen a case where someone tried to have a VM server and physical tape library attached via iSCSI.
It simply will not work. My advice is to not even try it.

VM backup server and tape library attached to physical SN will work very well, just bear in mind what Preston wrote about the AFTD or DD device for networker server bootstrap and index backups.

-----Original Message-----
From: EMC Data Protection Q & A <EMC-DATAPROTECTION-L@LISTSERV.TEMPLE.EDU> On Behalf Of Preston de Guise
Sent: 08 May 2019 21:41
To: EMC-DATAPROTECTION-L@LISTSERV.TEMPLE.EDU
Subject: Re: [EMC-DataProtection-L] Advice: upgrading from 8.x to latest (18.x?)

The NetWorker server can be virtualised; however, bear in mind it still needs to have a backup device to write to for bootstrap backups: in the case of a virtualised NetWorker server this should be an AFTD or Data Domain Boost device. Passthrough of tape to virtual machines isn't ever recommended. Even VMware themselves are pretty leery of it.

Cheers,
Preston.

Sent from my iPad

> On 9 May 2019, at 04:42, Tim Mooney <Tim.Mooney@ndsu.edu> wrote:
>
> In regard to: Re: [EMC-DataProtection-L] Advice: upgrading from 8.x to...:
>
>> What of the others points of consideration - can you have/is it
>> recommended to have the NW server itself be a VM? In my case, since I
>> need to keep writing to tape, and my tape library won't work via
>> iSCSI, I need to keep it connected via FC. Which means (I think) I
>> need to make a storage node that is FC connected, and define it to
>> use the library. And therefore change all my clone jobs to write to that library.
>
> We're not doing that here, but I'm pretty sure that a physical storage
> node + a VM NetWorker server is a supported configuration. As long as
> you've designed your VM environment with the necessary I/O bandwidth
> for indexes and the other I/O needs of a server, you should be fine.
>
> Tim
> --
> Tim Mooney Tim.Mooney@ndsu.edu
> Enterprise Computing & Infrastructure 701-231-1076 (Voice)
> Room 242-J6, Quentin Burdick Building 701-231-8541 (Fax)
> North Dakota State University, Fargo, ND 58105-5164
>
>
> --
> This list is hosted as a public service at Temple University by Stan
> Horwitz If you wish to sign off this list or adjust your subscription
> settings, please do so via
> http://listserv.temple.edu/archives/emc-dataprotection-l.html
> If you have any questions regarding management of this list, please
> send email to owner-emc-dataprotection-l@listserv.temple.edu


--
This list is hosted as a public service at Temple University by Stan Horwitz If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu


--
This list is hosted as a public service at Temple University by Stan Horwitz
If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu
This message was imported via the External PhorumMail Module
Tossavainen, Atro (Consultant)
Re: Advice: upgrading from 8.x to latest (18.x?)
May 16, 2019 01:59AM
Does anybody have any clues for improving tape cloning?


Our scenario is as Andy Fahy described - VM main server, physical SN, tape library attached to SN over FC, all disk backups made to DD.


As far as I have been able to determine cloning from DD to tape always goes via the main server. This makes the networking of the main server VM the bottleneck. We'd love to be able to take that out of the equation without resorting to explicit two-stage cloning where a bit of local disk on the SN acts as intermediate storage.


--
Atro Tossavainen
System Administrator (Consultant)
Oy IGT Finland Ab, a fully owned subsidiary of IGT plc (Business ID FI08945519)https://tietopalvelu.ytj.fi/yritystiedot.aspx?yavain=732317&tarkiste=0086DB0700906F9BA97F0CB46E633AADEC57A6B8
tel. +358-44-5000-600 (cell)
________________________________
From: EMC Data Protection Q & A <EMC-DATAPROTECTION-L@LISTSERV.TEMPLE.EDU> on behalf of Andy Fahy <farske10@HOTMAIL.COM>
Sent: Wednesday, May 15, 2019 11:20:24 PM
To: EMC-DATAPROTECTION-L@LISTSERV.TEMPLE.EDU
Subject: Re: [EMC-DataProtection-L] Advice: upgrading from 8.x to latest (18.x?)

[THIS MESSAGE ORIGINATED FROM A NON-IGT EMAIL ADDRESS]



I have seen a case where someone tried to have a VM server and physical tape library attached via iSCSI.
It simply will not work. My advice is to not even try it.

VM backup server and tape library attached to physical SN will work very well, just bear in mind what Preston wrote about the AFTD or DD device for networker server bootstrap and index backups.

-----Original Message-----
From: EMC Data Protection Q & A <EMC-DATAPROTECTION-L@LISTSERV.TEMPLE.EDU> On Behalf Of Preston de Guise
Sent: 08 May 2019 21:41
To: EMC-DATAPROTECTION-L@LISTSERV.TEMPLE.EDU
Subject: Re: [EMC-DataProtection-L] Advice: upgrading from 8.x to latest (18.x?)

The NetWorker server can be virtualised; however, bear in mind it still needs to have a backup device to write to for bootstrap backups: in the case of a virtualised NetWorker server this should be an AFTD or Data Domain Boost device. Passthrough of tape to virtual machines isn't ever recommended. Even VMware themselves are pretty leery of it.

Cheers,
Preston.

Sent from my iPad

> On 9 May 2019, at 04:42, Tim Mooney <Tim.Mooney@ndsu.edu> wrote:
>
> In regard to: Re: [EMC-DataProtection-L] Advice: upgrading from 8.x to...:
>
>> What of the others points of consideration - can you have/is it
>> recommended to have the NW server itself be a VM? In my case, since I
>> need to keep writing to tape, and my tape library won't work via
>> iSCSI, I need to keep it connected via FC. Which means (I think) I
>> need to make a storage node that is FC connected, and define it to
>> use the library. And therefore change all my clone jobs to write to that library.
>
> We're not doing that here, but I'm pretty sure that a physical storage
> node + a VM NetWorker server is a supported configuration. As long as
> you've designed your VM environment with the necessary I/O bandwidth
> for indexes and the other I/O needs of a server, you should be fine.
>
> Tim
> --
> Tim Mooney Tim.Mooney@ndsu.edu
> Enterprise Computing & Infrastructure 701-231-1076 (Voice)
> Room 242-J6, Quentin Burdick Building 701-231-8541 (Fax)
> North Dakota State University, Fargo, ND 58105-5164
>
>
> --
> This list is hosted as a public service at Temple University by Stan
> Horwitz If you wish to sign off this list or adjust your subscription
> settings, please do so via
> http://listserv.temple.edu/archives/emc-dataprotection-l.html
> If you have any questions regarding management of this list, please
> send email to owner-emc-dataprotection-l@listserv.temple.edu


--
This list is hosted as a public service at Temple University by Stan Horwitz If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu


--
This list is hosted as a public service at Temple University by Stan Horwitz
If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu
CONFIDENTIALITY NOTICE: This message is the property of International Game Technology PLC and/or its subsidiaries and may contain proprietary, confidential or trade secret information. This message is intended solely for the use of the addressee. If you are not the intended recipient and have received this message in error, please delete this message from your system. Any unauthorized reading, distribution, copying, or other use of this message or its attachments is strictly prohibited.


--
This list is hosted as a public service at Temple University by Stan Horwitz
If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu
This message was imported via the External PhorumMail Module
Preston de Guise
Re: Advice: upgrading from 8.x to latest (18.x?)
May 16, 2019 03:59AM
Hi Atro,

I’m assuming from what you’ve described that the DD devices are mounted on the backup server only?

Data Domain Boost devices (I’m assuming you’re talking Boost) can be simultaneously mounted on multiple storage nodes.

Have the Storage Node also mount the volumes. It should be able to then clone locally without reading through the NetWorker server.

Cheers,
Preston.


> On 16 May 2019, at 6:15 pm, Tossavainen, Atro (Consultant) <Atro.Tossavainen@IGT.COM> wrote:
>
> Does anybody have any clues for improving tape cloning?
>
> Our scenario is as Andy Fahy described - VM main server, physical SN, tape library attached to SN over FC, all disk backups made to DD.
>
> As far as I have been able to determine cloning from DD to tape always goes via the main server. This makes the networking of the main server VM the bottleneck. We'd love to be able to take that out of the equation without resorting to explicit two-stage cloning where a bit of local disk on the SN acts as intermediate storage.
>
> --
> Atro Tossavainen
> System Administrator (Consultant)
> Oy IGT Finland Ab, a fully owned subsidiary of IGT plc (Business ID FI08945519) https://tietopalvelu.ytj.fi/yritystiedot.aspx?yavain=732317&tarkiste=0086DB0700906F9BA97F0CB46E633AADEC57A6B8
> tel. +358-44-5000-600 (cell)
> From: EMC Data Protection Q & A <EMC-DATAPROTECTION-L@LISTSERV.TEMPLE.EDU <mailto:EMC-DATAPROTECTION-L@LISTSERV.TEMPLE.EDU>> on behalf of Andy Fahy <farske10@HOTMAIL.COM <mailto:farske10@HOTMAIL.COM>>
> Sent: Wednesday, May 15, 2019 11:20:24 PM
> To: EMC-DATAPROTECTION-L@LISTSERV.TEMPLE.EDU <mailto:EMC-DATAPROTECTION-L@LISTSERV.TEMPLE.EDU>
> Subject: Re: [EMC-DataProtection-L] Advice: upgrading from 8.x to latest (18.x?)
>
> [THIS MESSAGE ORIGINATED FROM A NON-IGT EMAIL ADDRESS]
>
>
>
> I have seen a case where someone tried to have a VM server and physical tape library attached via iSCSI.
> It simply will not work. My advice is to not even try it.
>
> VM backup server and tape library attached to physical SN will work very well, just bear in mind what Preston wrote about the AFTD or DD device for networker server bootstrap and index backups.
>
> -----Original Message-----
> From: EMC Data Protection Q & A <EMC-DATAPROTECTION-L@LISTSERV.TEMPLE.EDU <mailto:EMC-DATAPROTECTION-L@LISTSERV.TEMPLE.EDU>> On Behalf Of Preston de Guise
> Sent: 08 May 2019 21:41
> To: EMC-DATAPROTECTION-L@LISTSERV.TEMPLE.EDU <mailto:EMC-DATAPROTECTION-L@LISTSERV.TEMPLE.EDU>
> Subject: Re: [EMC-DataProtection-L] Advice: upgrading from 8.x to latest (18.x?)
>
> The NetWorker server can be virtualised; however, bear in mind it still needs to have a backup device to write to for bootstrap backups: in the case of a virtualised NetWorker server this should be an AFTD or Data Domain Boost device. Passthrough of tape to virtual machines isn't ever recommended. Even VMware themselves are pretty leery of it.
>
> Cheers,
> Preston.
>
> Sent from my iPad
>
> > On 9 May 2019, at 04:42, Tim Mooney <Tim.Mooney@ndsu.edu <mailto:Tim.Mooney@ndsu.edu>> wrote:
> >
> > In regard to: Re: [EMC-DataProtection-L] Advice: upgrading from 8.x to...:
> >
> >> What of the others points of consideration - can you have/is it
> >> recommended to have the NW server itself be a VM? In my case, since I
> >> need to keep writing to tape, and my tape library won't work via
> >> iSCSI, I need to keep it connected via FC. Which means (I think) I
> >> need to make a storage node that is FC connected, and define it to
> >> use the library. And therefore change all my clone jobs to write to that library.
> >
> > We're not doing that here, but I'm pretty sure that a physical storage
> > node + a VM NetWorker server is a supported configuration. As long as
> > you've designed your VM environment with the necessary I/O bandwidth
> > for indexes and the other I/O needs of a server, you should be fine.
> >
> > Tim
> > --
> > Tim Mooney Tim.Mooney@ndsu.edu <mailto:Tim.Mooney@ndsu.edu>
> > Enterprise Computing & Infrastructure 701-231-1076 (Voice)
> > Room 242-J6, Quentin Burdick Building 701-231-8541 (Fax)
> > North Dakota State University, Fargo, ND 58105-5164
> >
> >
> > --
> > This list is hosted as a public service at Temple University by Stan
> > Horwitz If you wish to sign off this list or adjust your subscription
> > settings, please do so via
> > http://listserv.temple.edu/archives/emc-dataprotection-l.html http://listserv.temple.edu/archives/emc-dataprotection-l.html
> > If you have any questions regarding management of this list, please
> > send email to owner-emc-dataprotection-l@listserv.temple.edu <mailto:owner-emc-dataprotection-l@listserv.temple.edu>
>
>
> --
> This list is hosted as a public service at Temple University by Stan Horwitz If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html http://listserv.temple.edu/archives/emc-dataprotection-l.html
> If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu <mailto:owner-emc-dataprotection-l@listserv.temple.edu>
>
>
> --
> This list is hosted as a public service at Temple University by Stan Horwitz
> If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html http://listserv.temple.edu/archives/emc-dataprotection-l.html
> If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu <mailto:owner-emc-dataprotection-l@listserv.temple.edu>
> CONFIDENTIALITY NOTICE: This message is the property of International Game Technology PLC and/or its subsidiaries and may contain proprietary, confidential or trade secret information. This message is intended solely for the use of the addressee. If you are not the intended recipient and have received this message in error, please delete this message from your system. Any unauthorized reading, distribution, copying, or other use of this message or its attachments is strictly prohibited.
> -- This list is hosted as a public service at Temple University by Stan Horwitz If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html http://listserv.temple.edu/archives/emc-dataprotection-l.html If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu <mailto:owner-emc-dataprotection-l@listserv.temple.edu>


--
This list is hosted as a public service at Temple University by Stan Horwitz
If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu
This message was imported via the External PhorumMail Module
Re: Advice: upgrading from 8.x to latest (18.x?)
May 16, 2019 04:28AM
Hi Atro,

until recently, we have attached 2 SNodes with 6 LTO5's each to clone the DD save sets.
Then we used scripted cloning to start cloning - one script for each DDBoost device.

This provided really enough throughput. Newer tape drive technology is even faster.
Michael Leone
Re: Advice: upgrading from 8.x to latest (18.x?)
May 16, 2019 06:59AM

This message was imported via the External PhorumMail Module
Matthew Powell
Re: Advice: upgrading from 8.x to latest (18.x?)
May 16, 2019 07:59AM
Good morning fellow backup admins,
Here at Clemson University we have solved the cloning process by having our Networker servers (all vm's) mount a SAM/QFS file system that has about 200 TB of online disk space. We have two SL8500 library with about 5 PB of tape storage although not all of that is for Legato. Most of it is for the HPC clusters we have here. The beauty of SAM/QFS is that when a file(s) land there for 30 minutes or 30 GB is reached then SAM/QFS makes two tape copies for me. One at the local site and the other (SL8500) is offsite so we have two copies made in one pass so cloning has been taken out of the equation. Networker see's only the 200 TB disk pool as advanced file type devices. If the 200 TB gets full on the SAM/QFS file system then it wipes the disk space since it has all files save to tape and then we have more disk space to use. It is quite an easy set up but without the HPC clusters we have here then the SL8500's can be quite expensive. Hope everyone has a wonderful day.

From: EMC Data Protection Q & A <EMC-DATAPROTECTION-L@LISTSERV.TEMPLE.EDU> On Behalf Of Michael Leone
Sent: Thursday, May 16, 2019 9:16 AM
To: EMC-DATAPROTECTION-L@LISTSERV.TEMPLE.EDU
Subject: Re: [EMC-DataProtection-L] Advice: upgrading from 8.x to latest (18.x?)

>I'm assuming from what you've described that the DD devices are
>mounted on the backup server only?
>
>
>Data Domain Boost devices (I'm assuming you're talking Boost) can be
>simultaneously mounted on multiple storage nodes.
>
>
>Have the Storage Node also mount the volumes. It should be able to
>then clone locally without reading through the NetWorker server.


Now, THAT is good to know! That configuration described is exactly what I want to go to - NW server as VM, all backups to DD (w/Boost); physical storage node, attached to tape library via FC.

Jobs save to DD. Then SN clones (hopefully directly from DD) to tape library ...


Thank you so much. Now I have both a destination, and a possible path to that destination ....

First find some hardware, and configure its FC to see the tape library.
Then make it a storage node, then mount the DD as drives there.
Reconfigure clone jobs to run through the storage node.

(I have to delete the library in NW, and re-add it, to the SN? I think so, right?)


Then, once that new configuration is stable, all that's left to decide:

do I try and upgrade the server from NW 8.x to latest (is it 18.x? Is it 9.x? I can barely tell ..) *first*, THEN virtualize it.

OR - virtualize it first, as a NW 8.x server. THEN upgrade it.

Either way, as a VM, expanding the index drive will be a lot easier than it will be as a physical box.

Thanks so much!



>
>
>Cheers,
>Preston.
>
>
>
>



--
This list is hosted as a public service at Temple University by Stan Horwitz
If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu
This message was imported via the External PhorumMail Module
Preston de Guise
Re: Advice: upgrading from 8.x to latest (18.x?)
May 17, 2019 01:59AM
Hi Atro,

Great to hear that this has made such a big difference for you! That’s exactly the sort of results and configuration flexibility I love to hear about.

Cheers!
Preston.

> On 17 May 2019, at 5:58 pm, Tossavainen, Atro (Consultant) <Atro.Tossavainen@IGT.com> wrote:
>
> I need to thank you Preston once more, even more.
>
> Prior to this, in order to avoid shoe-shining, we were operating a kludge where a two-stage cloning process would first take stuff from DD to temporary local disk space on the SN, then the local disk would be emptied onto tape. This prevented the mechanical problems, but took more time anyway, and things needed to be done in chunks of however much local disk space there was, which was inconvenient, and when an individual saveset exceeded the size of the temp pool, couldn't be done. Now that the SN mounts the DDs directly, we're seeing speeds to tape that even exceed the speed we saw from local disk, and all kludges ought to be history. Hooray!
>
> --
> Atro Tossavainen
> System Administrator (Consultant)
> Oy IGT Finland Ab, a fully owned subsidiary of IGT plc (Business ID FI08945519) https://tietopalvelu.ytj.fi/yritystiedot.aspx?yavain=732317&tarkiste=0086DB0700906F9BA97F0CB46E633AADEC57A6B8
> tel. +358-44-5000-600 (cell)
> From: Preston de Guise <enterprise.backup@gmail.com <mailto:enterprise.backup@gmail.com>>
> Sent: Thursday, May 16, 2019 1:27:23 PM
> To: EMC Data Protection Q & A; Tossavainen, Atro (Consultant)
> Subject: Re: [EMC-DataProtection-L] Advice: upgrading from 8.x to latest (18.x?)
>
> [THIS MESSAGE ORIGINATED FROM A NON-IGT EMAIL ADDRESS]
>
> Hi Atro,
>
> I’m assuming from what you’ve described that the DD devices are mounted on the backup server only?
>
> Data Domain Boost devices (I’m assuming you’re talking Boost) can be simultaneously mounted on multiple storage nodes.
>
> Have the Storage Node also mount the volumes. It should be able to then clone locally without reading through the NetWorker server.
>
> Cheers,
> Preston.
>
>
>> On 16 May 2019, at 6:15 pm, Tossavainen, Atro (Consultant) <Atro.Tossavainen@IGT.COM <mailto:Atro.Tossavainen@IGT.COM>> wrote:
>>
>> Does anybody have any clues for improving tape cloning?
>>
>> Our scenario is as Andy Fahy described - VM main server, physical SN, tape library attached to SN over FC, all disk backups made to DD.
>>
>> As far as I have been able to determine cloning from DD to tape always goes via the main server. This makes the networking of the main server VM the bottleneck. We'd love to be able to take that out of the equation without resorting to explicit two-stage cloning where a bit of local disk on the SN acts as intermediate storage.
>>
>> --
>> Atro Tossavainen
>> System Administrator (Consultant)
>> Oy IGT Finland Ab, a fully owned subsidiary of IGT plc (Business ID FI08945519) https://tietopalvelu.ytj.fi/yritystiedot.aspx?yavain=732317&tarkiste=0086DB0700906F9BA97F0CB46E633AADEC57A6B8
>> tel. +358-44-5000-600 (cell)



--
This list is hosted as a public service at Temple University by Stan Horwitz
If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu
This message was imported via the External PhorumMail Module
Tossavainen, Atro (Consultant)
Re: Advice: upgrading from 8.x to latest (18.x?)
May 17, 2019 05:00AM
I have to report one more interesting side effect.


I didn't have the old DD Boost user password saved away anywhere, so I figured I'll simply change the password and change the existing resources as well.


Not so fast. Something in the order I did things in was wrong and while cloning worked like a charm, new saves were failing because the old DD Boost user password was hoarded somewhere. Restarting NetWorker services made it go away, so no biggie, but a surprise.


--
Atro Tossavainen
System Administrator (Consultant)
Oy IGT Finland Ab, a fully owned subsidiary of IGT plc (Business ID FI08945519)https://tietopalvelu.ytj.fi/yritystiedot.aspx?yavain=732317&tarkiste=0086DB0700906F9BA97F0CB46E633AADEC57A6B8
tel. +358-44-5000-600 (cell)
________________________________
From: Preston de Guise <enterprise.backup@gmail.com>
Sent: Friday, May 17, 2019 11:18:23 AM
To: Tossavainen, Atro (Consultant)
Cc: EMC Data Protection Q & A
Subject: Re: [EMC-DataProtection-L] Advice: upgrading from 8.x to latest (18.x?)


[THIS MESSAGE ORIGINATED FROM A NON-IGT EMAIL ADDRESS]



Hi Atro,

Great to hear that this has made such a big difference for you! That’s exactly the sort of results and configuration flexibility I love to hear about.

Cheers!
Preston.

On 17 May 2019, at 5:58 pm, Tossavainen, Atro (Consultant) <Atro.Tossavainen@IGT.com<mailto:Atro.Tossavainen@IGT.com>> wrote:

I need to thank you Preston once more, even more.

Prior to this, in order to avoid shoe-shining, we were operating a kludge where a two-stage cloning process would first take stuff from DD to temporary local disk space on the SN, then the local disk would be emptied onto tape. This prevented the mechanical problems, but took more time anyway, and things needed to be done in chunks of however much local disk space there was, which was inconvenient, and when an individual saveset exceeded the size of the temp pool, couldn't be done. Now that the SN mounts the DDs directly, we're seeing speeds to tape that even exceed the speed we saw from local disk, and all kludges ought to be history. Hooray!

--
Atro Tossavainen
System Administrator (Consultant)
Oy IGT Finland Ab, a fully owned subsidiary of IGT plc (Business ID FI08945519)https://tietopalvelu.ytj.fi/yritystiedot.aspx?yavain=732317&tarkiste=0086DB0700906F9BA97F0CB46E633AADEC57A6B8
tel. +358-44-5000-600 (cell)
________________________________
From: Preston de Guise <enterprise.backup@gmail.com<mailto:enterprise.backup@gmail.com>>
Sent: Thursday, May 16, 2019 1:27:23 PM
To: EMC Data Protection Q & A; Tossavainen, Atro (Consultant)
Subject: Re: [EMC-DataProtection-L] Advice: upgrading from 8.x to latest (18.x?)

[THIS MESSAGE ORIGINATED FROM A NON-IGT EMAIL ADDRESS]



Hi Atro,

I’m assuming from what you’ve described that the DD devices are mounted on the backup server only?

Data Domain Boost devices (I’m assuming you’re talking Boost) can be simultaneously mounted on multiple storage nodes.

Have the Storage Node also mount the volumes. It should be able to then clone locally without reading through the NetWorker server.

Cheers,
Preston.


On 16 May 2019, at 6:15 pm, Tossavainen, Atro (Consultant) <Atro.Tossavainen@IGT.COM<mailto:Atro.Tossavainen@IGT.COM>> wrote:

Does anybody have any clues for improving tape cloning?

Our scenario is as Andy Fahy described - VM main server, physical SN, tape library attached to SN over FC, all disk backups made to DD.

As far as I have been able to determine cloning from DD to tape always goes via the main server. This makes the networking of the main server VM the bottleneck. We'd love to be able to take that out of the equation without resorting to explicit two-stage cloning where a bit of local disk on the SN acts as intermediate storage.

--
Atro Tossavainen
System Administrator (Consultant)
Oy IGT Finland Ab, a fully owned subsidiary of IGT plc (Business ID FI08945519)https://tietopalvelu.ytj.fi/yritystiedot.aspx?yavain=732317&tarkiste=0086DB0700906F9BA97F0CB46E633AADEC57A6B8
tel. +358-44-5000-600 (cell)

CONFIDENTIALITY NOTICE: This message is the property of International Game Technology PLC and/or its subsidiaries and may contain proprietary, confidential or trade secret information. This message is intended solely for the use of the addressee. If you are not the intended recipient and have received this message in error, please delete this message from your system. Any unauthorized reading, distribution, copying, or other use of this message or its attachments is strictly prohibited.


--
This list is hosted as a public service at Temple University by Stan Horwitz
If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu
This message was imported via the External PhorumMail Module
Preston de Guise
Re: Advice: upgrading from 8.x to latest (18.x?)
May 17, 2019 05:59PM
Hi Michael,

Sorry for the delayed response. Answers in-line.

> On 16 May 2019, at 11:16 pm, Michael Leone <Michael.Leone@PHA.PHILA.GOV> wrote:
>
> >I’m assuming from what you’ve described that the DD devices are
> >mounted on the backup server only?
> >
> >
> >Data Domain Boost devices (I’m assuming you’re talking Boost) can be
> >simultaneously mounted on multiple storage nodes.
> >
> >
> >Have the Storage Node also mount the volumes. It should be able to
> >then clone locally without reading through the NetWorker server.
>
>
> Now, THAT is good to know! That configuration described is exactly what I want to go to - NW server as VM, all backups to DD (w/Boost); physical storage node, attached to tape library via FC.
>
> Jobs save to DD. Then SN clones (hopefully directly from DD) to tape library ...
>
>
> Thank you so much. Now I have both a destination, and a possible path to that destination ....
>
> First find some hardware, and configure its FC to see the tape library.
> Then make it a storage node, then mount the DD as drives there.
> Reconfigure clone jobs to run through the storage node.
>
> (I have to delete the library in NW, and re-add it, to the SN? I think so, right?)

Yes, if you’re moving the library from the server to a storage node, you’ll need to delete and re-add it. (If you’re just changing the SCSI address the Library is seen at on the same host, you can simply disable the library, edit its config and adjust that path, then re-enable the library.)

> Then, once that new configuration is stable, all that's left to decide:
>
> do I try and upgrade the server from NW 8.x to latest (is it 18.x? Is it 9.x? I can barely tell ..) *first*, THEN virtualize it.
>
> OR - virtualize it first, as a NW 8.x server. THEN upgrade it.
>
> Either way, as a VM, expanding the index drive will be a lot easier than it will be as a physical box.

If you’re on NetWorker v8.x, I’d probably be inclined to upgrade NetWorker first, then virtualize the NetWorker server, simply because then the NetWorker services are fully supported while you’re going through the process. And yes, index drive management is a lot easier when the NetWorker server is a VM!

I’d recommend 18.2.x. but make sure you review the upgrade guide. I don’t 100% recall, but I have a sneaking suspicion you need to go from 8 to 18.x via a temporary 9.x upgrade path. The upgrade guide will confirm one way or the other though.

Cheers,
Preston.



--
This list is hosted as a public service at Temple University by Stan Horwitz
If you wish to sign off this list or adjust your subscription settings, please do so via http://listserv.temple.edu/archives/emc-dataprotection-l.html
If you have any questions regarding management of this list, please send email to owner-emc-dataprotection-l@listserv.temple.edu
This message was imported via the External PhorumMail Module
Sorry, only registered users may post in this forum.

Click here to login