Have an all netware cluster right now. It is running the following services.
Groupwise POA, MTA's Gwia Webacc
Gw messesnger
iprint
file services
dns
dhcp

The goal is to move everything from NW to OES2 linux except gw stuff. Is
there a safe way to do this that the cluster will be okay and be a mixed
environment. Gw will only migrate between nw servers. Rest willmigrate on
linux servers. I have read that could have issues creating pools and
volumes after doing this.

Thanks

Just because the data is all held on the one SAN, doesn't mean that
you cannot split it into two. I assume that you have 3 or more nodes
on the NW side and want to keep three as a minimum for the GW. The
OES2 would also require 3 (2 isn't a cluster :-)

You could build the new cluster, connected to the same SAN, get the
new iPrint etc working and tested - same with DNS and DHCP and then
when the time comes, to move the file - offline the resource on the NW
cluster - change the zoning / masking so that the Linux servers can
now see the resource - use NSSMU to remove the "shared for clustering"
attribute - mount the volume and update eDir on one of the Linux
nodes, pull the object into eDir and then cluster it... Moved!

make sense?



On Wed, 28 May 2008 23:45:56 GMT, Rick Bousquet
<rdbat@rdbnetsolutionsdot.com> wrote:

>Have an all netware cluster right now. It is running the following services.
>Groupwise POA, MTA's Gwia Webacc
>Gw messesnger
>iprint
>file services
>dns
>dhcp
>
>The goal is to move everything from NW to OES2 linux except gw stuff. Is
>there a safe way to do this that the cluster will be okay and be a mixed
>environment. Gw will only migrate between nw servers. Rest willmigrate on
>linux servers. I have read that could have issues creating pools and
>volumes after doing this.
>
>Thanks
>
Tim
___________________
Tim Heywood (SYSOP)
NDS8
Scotland
(God's Country)
www.nds8.co.uk
___________________

In theory, practice and theory are the same
In Practice, they are different

Tim Heywood NSC SYSOP wrote:

> Just because the data is all held on the one SAN, doesn't mean that
> you cannot split it into two. I assume that you have 3 or more nodes
> on the NW side and want to keep three as a minimum for the GW. The
> OES2 would also require 3 (2 isn't a cluster :-)
>
> You could build the new cluster, connected to the same SAN, get the
> new iPrint etc working and tested - same with DNS and DHCP and then
> when the time comes, to move the file - offline the resource on the NW
> cluster - change the zoning / masking so that the Linux servers can
> now see the resource - use NSSMU to remove the "shared for clustering"
> attribute - mount the volume and update eDir on one of the Linux
> nodes, pull the object into eDir and then cluster it... Moved!
>
> make sense?
>
>
>
> On Wed, 28 May 2008 23:45:56 GMT, Rick Bousquet
> <rdbat@rdbnetsolutionsdot.com> wrote:
>
>>Have an all netware cluster right now. It is running the following
>>services. Groupwise POA, MTA's Gwia Webacc
>>Gw messesnger
>>iprint
>>file services
>>dns
>>dhcp
>>
>>The goal is to move everything from NW to OES2 linux except gw stuff. Is
>>there a safe way to do this that the cluster will be okay and be a mixed
>>environment. Gw will only migrate between nw servers. Rest willmigrate on
>>linux servers. I have read that could have issues creating pools and
>>volumes after doing this.
>>
>>Thanks
>>
> Tim
> ___________________
> Tim Heywood (SYSOP)
> NDS8
> Scotland
> (God's Country)
> www.nds8.co.uk
> ___________________
>
> In theory, practice and theory are the same
> In Practice, they are different

The biggest issue to over come I know will be the san itself. The customer
actually has an apple xserve raid. It works well but had a lot of
limitations how you could slice it up. That is why I am worried this will
be a backup and restore more than anything.

So that is presented as an iSCSI target? Should be OK - the problem
is normally the extra hardware required during the cut over.

On Thu, 29 May 2008 11:07:58 GMT, Rick Bousquet
<rdbat@rdbnetsolutionsdot.com> wrote:

>
>The biggest issue to over come I know will be the san itself. The customer
>actually has an apple xserve raid. It works well but had a lot of
>limitations how you could slice it up. That is why I am worried this will
>be a backup and restore more than anything.
Tim
___________________
Tim Heywood (SYSOP)
NDS8
Scotland
(God's Country)
www.nds8.co.uk
___________________

In theory, practice and theory are the same
In Practice, they are different

Tim Heywood NSC SYSOP wrote:

> So that is presented as an iSCSI target? Should be OK - the problem
> is normally the extra hardware required during the cut over.
>
> On Thu, 29 May 2008 11:07:58 GMT, Rick Bousquet
> <rdbat@rdbnetsolutionsdot.com> wrote:
>
>>
>>The biggest issue to over come I know will be the san itself. The customer
>>actually has an apple xserve raid. It works well but had a lot of
>>limitations how you could slice it up. That is why I am worried this will
>>be a backup and restore more than anything.
> Tim
> ___________________
> Tim Heywood (SYSOP)
> NDS8
> Scotland
> (God's Country)
> www.nds8.co.uk
> ___________________
>
> In theory, practice and theory are the same
> In Practice, they are different

No it is a fiber channel san.

I wonder who they re badge - if the storage is over fibre channel then
there shouldn't be a problem.

usual 1-1-1-1 lun-partition-pool-volume rules apply.






On Thu, 29 May 2008 21:00:17 GMT, Rick Bousquet
<rdbat@rdbnetsolutionsdot.com> wrote:

>
>No it is a fiber channel san.
Tim
___________________
Tim Heywood (SYSOP)
NDS8
Scotland
(God's Country)
www.nds8.co.uk
___________________

In theory, practice and theory are the same
In Practice, they are different