Hi,
we recently updated ONTAP from 7.2.5.1 to 7.3 on our 3140c, which led to many problems : one head doesn't boot anymore, so the cluster failover is impossible; the other head (currently in takeover) seems to reboot randomly every week; and when this happens, it freezes some of our RHEL4 servers (kernel panics, somewhere in lockd).
Putted shortly: this is a mess.
The issues are currently investigated by NetApp, but it may take some time until we see a corrective action on their side. So we're investigating another solution, which would be to downgrade back into 7.2.5.1 (since it appears 7.2.6 doesn't seem very stable either).
Would any of you know how to do that, or could point me at a proper procedure somewhere? NetApp support was unresponsive on this one, and I'm a bit concerned about possible changes to the WAFL layout or something similar between 7.2 and 7.3.
Simon
Hi
You have probably already guessed that this procedure is not for the faint-hearted. The reversion of the WAFL file system from 7.3 to 7.2 will involve some downtime and hence cause some pain.
The first step will be to sort out what is wrong with the cluster, because you will want to revert both heads in the cluster. This is something you'd want to tackle with support on the line.
Having done that, then the steps will look somewhat like this:
- Download the system files for 7.2.5.1 from now.netapp.com. Be sure to download the system files that match your node model for a Windows host. Copy the .exe file to /etc/software using either the CIFS ETC$ share or the NFS /vol/vol0/etc export. - In order to revert you will need to switch off all the services such as iscsi, cifs, nfs and snapmirror. - cifs terminate -t 5 - nfs off - snapmirror off - iscsi stop - cf disable
- Then fire off an autosupport and install the software - options autosupport.doit Reverting - software install 7251_setup_q.exe - download
- Then start the revert process; this can last a long time: - revert_to 7.2
When reversion completes the filer will halt and so you will need to restart it by hand using the bye command at the firmware prompt.
I would also ensure you have a console connection whilst doing this to monitor the progress.
Bonne chance! Kindest regards Kenneth
Date: Tue, 25 Nov 2008 13:43:19 +0100 From: svallet@genoscope.cns.fr To: toasters@mathworks.com Subject: Downgrading Data ONTAP?
Hi,
we recently updated ONTAP from 7.2.5.1 to 7.3 on our 3140c, which led to many problems : one head doesn't boot anymore, so the cluster failover is impossible; the other head (currently in takeover) seems to reboot randomly every week; and when this happens, it freezes some of our RHEL4 servers (kernel panics, somewhere in lockd).
Putted shortly: this is a mess.
The issues are currently investigated by NetApp, but it may take some time until we see a corrective action on their side. So we're investigating another solution, which would be to downgrade back into 7.2.5.1 (since it appears 7.2.6 doesn't seem very stable either).
Would any of you know how to do that, or could point me at a proper procedure somewhere? NetApp support was unresponsive on this one, and I'm a bit concerned about possible changes to the WAFL layout or something similar between 7.2 and 7.3.
Simon
_________________________________________________________________ Express yourself instantly with MSN Messenger! Download today it's FREE! http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
On Tue, 25 Nov 2008 05:08:15 -0800 "Winkel, Paul" Paul.Winkel@netapp.com wrote:
Anyway, here is the link. http://now.netapp.com/NOW/knowledge/docs/ontap/rel73/html/ontap/upgrade/ upgrading/concept/c_oc_upg_revert.html
I don't know how I could possibly miss that one.
On Tue, 25 Nov 2008 14:13:55 +0100 Kenneth Heal kheal@hotmail.com wrote:
You have probably already guessed that this procedure is not for the faint-hearted. The reversion of the WAFL file system from 7.3 to 7.2 will involve some downtime and hence cause some pain.
Somehow I expected that :-(
Many thanks to both of you -- now I may need to read some docs.
Simon
Hi
I shall also add this link: http://now.netapp.com/NOW/knowledge/docs/ontap/rel73/html/ontap/upgrade/upgr...
This is the reversion section of the 7.3 upgrade manual and includes the known gotchas.
regards Kenneth
From: kheal@hotmail.com To: svallet@genoscope.cns.fr; toasters@mathworks.com Subject: RE: Downgrading Data ONTAP? Date: Tue, 25 Nov 2008 14:13:55 +0100
Hi
You have probably already guessed that this procedure is not for the faint-hearted. The reversion of the WAFL file system from 7.3 to 7.2 will involve some downtime and hence cause some pain.
The first step will be to sort out what is wrong with the cluster, because you will want to revert both heads in the cluster. This is something you'd want to tackle with support on the line.
Having done that, then the steps will look somewhat like this:
Download the system files for 7.2.5.1 from now.netapp.com. Be sure to download the system files that match your node model for a Windows host. Copy the .exe file to /etc/software using either the CIFS ETC$ share or the NFS /vol/vol0/etc export.
In order to revert you will need to switch off all the services such as iscsi, cifs, nfs and snapmirror.
- cifs terminate -t 5
- nfs off
- snapmirror off
- iscsi stop
- cf disable
Then fire off an autosupport and install the software
- options autosupport.doit Reverting
- software install 7251_setup_q.exe
- download
Then start the revert process; this can last a long time:
- revert_to 7.2
When reversion completes the filer will halt and so you will need to restart it by hand using the bye command at the firmware prompt.
I would also ensure you have a console connection whilst doing this to monitor the progress.
Bonne chance! Kindest regards Kenneth
Date: Tue, 25 Nov 2008 13:43:19 +0100 From: svallet@genoscope.cns.fr To: toasters@mathworks.com Subject: Downgrading Data ONTAP?
Hi,
we recently updated ONTAP from 7.2.5.1 to 7.3 on our 3140c, which led to many problems : one head doesn't boot anymore, so the cluster failover is impossible; the other head (currently in takeover) seems to reboot randomly every week; and when this happens, it freezes some of our RHEL4 servers (kernel panics, somewhere in lockd).
Putted shortly: this is a mess.
The issues are currently investigated by NetApp, but it may take some time until we see a corrective action on their side. So we're investigating another solution, which would be to downgrade back into 7.2.5.1 (since it appears 7.2.6 doesn't seem very stable either).
Would any of you know how to do that, or could point me at a proper procedure somewhere? NetApp support was unresponsive on this one, and I'm a bit concerned about possible changes to the WAFL layout or something similar between 7.2 and 7.3.
Simon
Express yourself instantly with MSN Messenger! Download today it's FREE! http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
_________________________________________________________________ Express yourself instantly with MSN Messenger! Download today it's FREE! http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
Dunno for sure if this also is valid for the upgrade to 7.3, but part of the revert_to process used to be deletion of snapshots which were made using the uprev ONTAP version.
This usually means you have to either delete all snapshots, or delete a few by hand, which can cause you to have to resync snapmirror/vaults.
Good luck on the reversion, and let us know how it went?
Greetings,
Nils
On Tue, Nov 25, 2008 at 3:17 PM, Kenneth Heal kheal@hotmail.com wrote:
Hi
I shall also add this link: http://now.netapp.com/NOW/knowledge/docs/ontap/rel73/html/ontap/upgrade/upgr...
This is the reversion section of the 7.3 upgrade manual and includes the known gotchas.
regards Kenneth
From: kheal@hotmail.com To: svallet@genoscope.cns.fr; toasters@mathworks.com Subject: RE: Downgrading Data ONTAP? Date: Tue, 25 Nov 2008 14:13:55 +0100
Hi
You have probably already guessed that this procedure is not for the faint-hearted. The reversion of the WAFL file system from 7.3 to 7.2 will involve some downtime and hence cause some pain.
The first step will be to sort out what is wrong with the cluster, because you will want to revert both heads in the cluster. This is something you'd want to tackle with support on the line.
Having done that, then the steps will look somewhat like this:
- Download the system files for 7.2.5.1 from now.netapp.com. Be sure to
download the system files that match your node model for a Windows host. Copy the .exe file to /etc/software using either the CIFS ETC$ share or the NFS /vol/vol0/etc export.
- In order to revert you will need to switch off all the services such as
iscsi, cifs, nfs and snapmirror.
cifs terminate -t 5
nfs off
snapmirror off
iscsi stop
cf disable
Then fire off an autosupport and install the software
options autosupport.doit Reverting
software install 7251_setup_q.exe
download
Then start the revert process; this can last a long time:
revert_to 7.2
When reversion completes the filer will halt and so you will need to restart it by hand using the bye command at the firmware prompt.
I would also ensure you have a console connection whilst doing this to monitor the progress.
Bonne chance! Kindest regards Kenneth
Date: Tue, 25 Nov 2008 13:43:19 +0100 From: svallet@genoscope.cns.fr To: toasters@mathworks.com Subject: Downgrading Data ONTAP?
Hi,
we recently updated ONTAP from 7.2.5.1 to 7.3 on our 3140c, which led to many problems : one head doesn't boot anymore, so the cluster failover is impossible; the other head (currently in takeover) seems to reboot randomly every week; and when this happens, it freezes some of our RHEL4 servers (kernel panics, somewhere in lockd).
Putted shortly: this is a mess.
The issues are currently investigated by NetApp, but it may take some time until we see a corrective action on their side. So we're investigating another solution, which would be to downgrade back into 7.2.5.1 (since it appears 7.2.6 doesn't seem very stable either).
Would any of you know how to do that, or could point me at a proper procedure somewhere? NetApp support was unresponsive on this one, and I'm a bit concerned about possible changes to the WAFL layout or something similar between 7.2 and 7.3.
Simon
Express yourself instantly with MSN Messenger! Download today it's FREE! http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
Express yourself instantly with MSN Messenger! MSN Messenger
Hi Nils,
On Tue, 25 Nov 2008 16:11:08 +0100 "Nils Vogels" bacardicoke@gmail.com wrote:
Dunno for sure if this also is valid for the upgrade to 7.3, but part of the revert_to process used to be deletion of snapshots which were made using the uprev ONTAP version.
This usually means you have to either delete all snapshots, or delete a few by hand, which can cause you to have to resync snapmirror/vaults.
Good luck on the reversion, and let us know how it went?
Thank you! FYI, the migration we discussed back in July (and which,for other reasons, was postponed to Nov, 8th) did go well.
We actually believe it is possible to transparently (without NFS service interruption) migrate data between filers without using virtualisation technologies, but this would require some modifications to the way snapmirror migrate functions. So we had to plan some downtime.
We wanted to suggest those to NetApp, but let's say we have some other problems for now ;-)
Simon
Hallo Nils Salut Simon
There is indeed a lot of changes in snapmirror (and hence also snapvault) code between 7.2 and 7.3 and so I suspect/fear you (Nils) may well be right. I am quite sure that any snapshots created post-upgrade will be killed in the process.
I am lucky to have not done this process 'in anger' (in a production environment) and have been able to confine it to lab testing.
Simon, let us know how it goes.
regards Kenneth
Date: Tue, 25 Nov 2008 16:11:08 +0100 From: bacardicoke@gmail.com To: toasters@mathworks.com Subject: Re: Downgrading Data ONTAP?
Dunno for sure if this also is valid for the upgrade to 7.3, but part of the revert_to process used to be deletion of snapshots which were made using the uprev ONTAP version.
This usually means you have to either delete all snapshots, or delete a few by hand, which can cause you to have to resync snapmirror/vaults.
Good luck on the reversion, and let us know how it went?
Greetings,
Nils
On Tue, Nov 25, 2008 at 3:17 PM, Kenneth Heal wrote:
Hi
I shall also add this link: http://now.netapp.com/NOW/knowledge/docs/ontap/rel73/html/ontap/upgrade/upgr...
This is the reversion section of the 7.3 upgrade manual and includes the known gotchas.
regards Kenneth
From: kheal@hotmail.com To: svallet@genoscope.cns.fr; toasters@mathworks.com Subject: RE: Downgrading Data ONTAP? Date: Tue, 25 Nov 2008 14:13:55 +0100
Hi
You have probably already guessed that this procedure is not for the faint-hearted. The reversion of the WAFL file system from 7.3 to 7.2 will involve some downtime and hence cause some pain.
The first step will be to sort out what is wrong with the cluster, because you will want to revert both heads in the cluster. This is something you'd want to tackle with support on the line.
Having done that, then the steps will look somewhat like this:
- Download the system files for 7.2.5.1 from now.netapp.com. Be sure to
download the system files that match your node model for a Windows host. Copy the .exe file to /etc/software using either the CIFS ETC$ share or the NFS /vol/vol0/etc export.
- In order to revert you will need to switch off all the services such as
iscsi, cifs, nfs and snapmirror.
cifs terminate -t 5
nfs off
snapmirror off
iscsi stop
cf disable
Then fire off an autosupport and install the software
options autosupport.doit Reverting
software install 7251_setup_q.exe
download
Then start the revert process; this can last a long time:
revert_to 7.2
When reversion completes the filer will halt and so you will need to restart it by hand using the bye command at the firmware prompt.
I would also ensure you have a console connection whilst doing this to monitor the progress.
Bonne chance! Kindest regards Kenneth
Date: Tue, 25 Nov 2008 13:43:19 +0100 From: svallet@genoscope.cns.fr To: toasters@mathworks.com Subject: Downgrading Data ONTAP?
Hi,
we recently updated ONTAP from 7.2.5.1 to 7.3 on our 3140c, which led to many problems : one head doesn't boot anymore, so the cluster failover is impossible; the other head (currently in takeover) seems to reboot randomly every week; and when this happens, it freezes some of our RHEL4 servers (kernel panics, somewhere in lockd).
Putted shortly: this is a mess.
The issues are currently investigated by NetApp, but it may take some time until we see a corrective action on their side. So we're investigating another solution, which would be to downgrade back into 7.2.5.1 (since it appears 7.2.6 doesn't seem very stable either).
Would any of you know how to do that, or could point me at a proper procedure somewhere? NetApp support was unresponsive on this one, and I'm a bit concerned about possible changes to the WAFL layout or something similar between 7.2 and 7.3.
Simon
Express yourself instantly with MSN Messenger! Download today it's FREE! http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
Express yourself instantly with MSN Messenger! MSN Messenger
-- Simple guidelines to happiness: Work like you don't need the money, Love like your heart has never been broken and Dance like no one can see you.
_________________________________________________________________ Express yourself instantly with MSN Messenger! Download today it's FREE! http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
-
Kenneth Heal -
Nils Vogels -
Simon Vallet