In any case, even if it does work, I would most certainly do an immediate takeover giveback. I would not even try it. I would shut the cluster down and recable to be safe.
************************************************* ** I do not support or endorse this method! ** ** If you care to try it is at your own risk ** ** I take no responsibility for your actions ** ** This is 100% unsupported ** ** Again, If I were doing this, ** ** I would shut down the cluster, ** ** recable and bring back up ** *************************************************
If you want to try while it is online, I might try this:
1. Do a takeover. (halt the node being taken over) 2. On surviving node, Disable one path to the shelf. 1. This leaves you vulnerable with a single path! 2. Do not touch any cables on surviving path! 3. Recable down path only on both the down node and the Up node 4. Bring the second path back up on the surviving node. 5. Bring the down node up. 1. You will likely NOT have Multi-path HA right now. 6. Wait about 8 minutes for protocol stabilization 7. Do a takeover in the other direction. 8. On surviving node, Disable one path to the shelf. 1. This leaves you vulnerable with a single path! 2. Do not touch any cables on surviving path! 9. Recable down path only on both the down node and the Up node 10. Bring the second path back up on the surviving node 11. Bring the down node up. 1. You will likely NOT have Multi-path HA right now. 12. Wait about 8 minutes for protocol stabilization 13. At this point the cables should be good. 14. Perform a takeover/giveback one more time in each direction.
--tmac
*Tim McCarthy, **Principal Consultant*
*Proud Member of the #NetAppATeam https://twitter.com/NetAppATeam*
*I Blog at TMACsRack https://tmacsrack.wordpress.com/*
On Thu, Jun 25, 2020 at 5:12 AM Heino Walther hw@beardmann.dk wrote:
Supported is one thing 😊 The question is if it panics, or keeps running…. Sadly I do not have a system to test this on…
/Heino
*Fra: *"andrei.borzenkov@fujitsu.com" andrei.borzenkov@fujitsu.com *Dato: *torsdag den 25. juni 2020 kl. 11.00 *Til: *Heino Walther hw@beardmann.dk, "toasters@teaparty.net" < toasters@teaparty.net> *Emne: *RE: Quick SAS Cable question
So you want to move shelf with data to different stack in the same filer, online? That was not supported officially last time I checked. Only empty shelves can be removed and reconnected to different stack.
Sent from my Fujitsu LIFEBOOK S937
With best regards
*Andrei Borzenkov*
Senior System Engineer
FJ EMEIA PR FOCP RU SM FSO
*FUJITSU*
Zemlyanoy val 9, 105064 Moscow, Russia
Tel.: +7 (495) 730 6220 ext. 2247
Mob.: +7 (916) 678 7208
E-mail: Andrei.Borzenkov@ts.fujitsu.com
Web: ts.fujitsu.com
Company details: OOO Fujitsu Technology Solutions / ts.fujitsu.com/imprint http://ts.fujitsu.com/imprint.html
This communication contains information that is confidential, proprietary in nature and/or privileged. It is for the exclusive use of the intended recipient(s). If you are not the intended recipient(s) or the person responsible for delivering it to the intended recipient(s), please note that any form of dissemination, distribution or copying of this communication is strictly prohibited and may be unlawful. If you have received this communication in error, please immediately notify the sender and delete the original communication. Thank you for your cooperation.
Please be advised that neither Fujitsu, its affiliates, its employees or agents accept liability for any errors, omissions or damages caused by delays of receipt or by any virus infection in this message or its attachments, or which may otherwise arise as a result of this e-mail transmission.
*From:* Heino Walther hw@beardmann.dk *Sent:* Thursday, June 25, 2020 11:18 AM *To:* Borzenkov, Andrei andrei.borzenkov@fujitsu.com; toasters@teaparty.net *Subject:* Re: Quick SAS Cable question
Hi Andrei
We have an aggregate spanning all three shelfs.
What I was thinking was to disconnect the A side of disk shelf (which is possible without any issues)
Then connect up the A side of the shelfs so that they now represent two loops, so two shelfs on one loop, and the last one on its own loop.
I will of cause make sure the controllers are able to see the disks before I move onto the B side of the shelfs…
I’m just afraid that ONTAP might be confused and panic..
/Heino
*Fra: *"andrei.borzenkov@fujitsu.com" andrei.borzenkov@fujitsu.com *Dato: *torsdag den 25. juni 2020 kl. 10.12 *Til: *Heino Walther hw@beardmann.dk, "toasters@teaparty.net" < toasters@teaparty.net> *Emne: *RE: Quick SAS Cable question
So you have three shelves and aggregate spanning these three shelves that you need to connect to FAS in two stacks, right? It is no problem offline (assuming you have the correct cables), when target FAS is switched off. Doing this online will result in incomplete aggregate unless you manage connecting both stacks almost at the same moment. Depending on raid groups layout, either aggregate won’t come online until you connect missing shelf or FAS may start rebuild due to missing disks.
If you mean something else or you have something different than FAS (there as E-Series etc), please provide more details.
Sent from my Fujitsu LIFEBOOK S937
With best regards
*Andrei Borzenkov*
Senior System Engineer
FJ EMEIA PR FOCP RU SM FSO
*FUJITSU*
Zemlyanoy val 9, 105064 Moscow, Russia
Tel.: +7 (495) 730 6220 ext. 2247
Mob.: +7 (916) 678 7208
E-mail: Andrei.Borzenkov@ts.fujitsu.com
Web: ts.fujitsu.com
Company details: OOO Fujitsu Technology Solutions / ts.fujitsu.com/imprint http://ts.fujitsu.com/imprint.html
This communication contains information that is confidential, proprietary in nature and/or privileged. It is for the exclusive use of the intended recipient(s). If you are not the intended recipient(s) or the person responsible for delivering it to the intended recipient(s), please note that any form of dissemination, distribution or copying of this communication is strictly prohibited and may be unlawful. If you have received this communication in error, please immediately notify the sender and delete the original communication. Thank you for your cooperation.
Please be advised that neither Fujitsu, its affiliates, its employees or agents accept liability for any errors, omissions or damages caused by delays of receipt or by any virus infection in this message or its attachments, or which may otherwise arise as a result of this e-mail transmission.
*From:* Toasters toasters-bounces@teaparty.net *On Behalf Of *Heino Walther *Sent:* Thursday, June 25, 2020 10:53 AM *To:* toasters@teaparty.net *Subject:* Quick SAS Cable question
Hi guys
I’m in the process of moving shelfs from one system to another, and has ran into a pickle where I would like to place one of three shelfs in a loop into its own loop.
Of cause the disks are a part of an aggregate and in production…
Can this be done without any issues? Just one path of the loop at a time?
I would imagine that it should work fine, but just checking… 😊
/Heino
Heino Walther https://www.linkedin.com/in/heinowalther/
Beardmann ApS http://beardmann.dk/
Jellingvej 9 - 7100 Vejle https://goo.gl/maps/xQVPFMHXpXu
D: 7199 9060 M: 2075 7501
--
Toasters mailing list Toasters@teaparty.net https://www.teaparty.net/mailman/listinfo/toasters
