cDOT cluster kernel-service

List overview All Threads
Download

newer

older

Berk Comm...

Re: Adding DS2246 to FAS2240-2...

Weber, Mark A

23 Apr 2015 23 Apr '15

5:26 p.m.

Anyone have a 4-node 2240 cluster running 8.2.x?

If so, could you tell me if failover says enabled or disabled when you do a "cluster kernel-service show" in diag mode. I've had a performance case open for months on this cluster and I noticed recently that this cluster says failover is disabled but our 4 node 3220 cluster says enabled.

LCNA2::> set d

Warning: These diagnostic commands are for use by NetApp personnel only. Do you want to continue? {y|n}: y

LCNA2::*> cluster kernel-service show Node State Ver Failover Dump FSM-State ---------- -------- ---- --------- --------- --------- LCNA2-01 enabled 1 disabled enabled CLAM_CORE_ST_NONMASTER_WAITING_FOR_BACKOFF_TIMEOUT LCNA2-02 enabled 1 disabled enabled CLAM_CORE_ST_NONMASTER_WAITING_FOR_BACKOFF_TIMEOUT LCNA2-03 enabled 1 disabled enabled CLAM_CORE_ST_MASTER_WAITING_FOR_HB_TIMEOUT LCNA2-04 enabled 1 disabled enabled CLAM_CORE_ST_NONMASTER_WAITING_FOR_BACKOFF_TIMEOUT 4 entries were displayed.

Thanks mark

Attachments:

attachment.html (text/html — 3.6 KB)

Show replies by date

Parisi, Justin

24 Apr 24 Apr

12:21 a.m.

That is for failover due to KCS being OOQ. Basically, if a node goes out of SAN quorum, it reboots.

What releases are your two clusters on? May just be a change in the default behaviors.

From: <Weber>, Mark A <mark-a-weber@uiowa.edumailto:mark-a-weber@uiowa.edu> Date: Thursday, April 23, 2015 at 1:26 PM To: "Toasters@teaparty.netmailto:Toasters@teaparty.net" <Toasters@teaparty.netmailto:Toasters@teaparty.net> Subject: cDOT cluster kernel-service

Anyone have a 4-node 2240 cluster running 8.2.x?

LCNA2::> set d

Warning: These diagnostic commands are for use by NetApp personnel only. Do you want to continue? {y|n}: y

Thanks mark

Weber, Mark A

12:47 a.m.

They are both 8.2.3P1 So that failover is for SAN only? Neither cluster has FC or iSCSI enabled

thanks mark

From: Parisi, Justin [mailto:Justin.Parisi@netapp.com] Sent: Thursday, April 23, 2015 7:22 PM To: Weber, Mark A; toasters@teaparty.net Subject: Re: cDOT cluster kernel-service

That is for failover due to KCS being OOQ. Basically, if a node goes out of SAN quorum, it reboots.

What releases are your two clusters on? May just be a change in the default behaviors.

Anyone have a 4-node 2240 cluster running 8.2.x?

LCNA2::> set d

Warning: These diagnostic commands are for use by NetApp personnel only. Do you want to continue? {y|n}: y

Thanks mark

Parisi, Justin

1:07 a.m.

Right. So if only NAS, disable it. It's only for SAN kernel quorum and automatic failover when OOQ. Storage failover (SFO) is different and is viewed via "storage failover show." You do want SFO enabled.

One of the clusters might have had it enabled as a holdover from a previous upgrade, or perhaps you had SAN enabled at one time.

From: <Weber>, Mark A <mark-a-weber@uiowa.edumailto:mark-a-weber@uiowa.edu> Date: Thursday, April 23, 2015 at 8:47 PM To: Justin Parisi <Justin.Parisi@netapp.commailto:Justin.Parisi@netapp.com>, "Toasters@teaparty.netmailto:Toasters@teaparty.net" <Toasters@teaparty.netmailto:Toasters@teaparty.net> Subject: RE: cDOT cluster kernel-service

They are both 8.2.3P1 So that failover is for SAN only? Neither cluster has FC or iSCSI enabled

thanks mark

From: Parisi, Justin [mailto:Justin.Parisi@netapp.com] Sent: Thursday, April 23, 2015 7:22 PM To: Weber, Mark A; toasters@teaparty.netmailto:toasters@teaparty.net Subject: Re: cDOT cluster kernel-service

That is for failover due to KCS being OOQ. Basically, if a node goes out of SAN quorum, it reboots.

What releases are your two clusters on? May just be a change in the default behaviors.

Anyone have a 4-node 2240 cluster running 8.2.x?

LCNA2::> set d

Warning: These diagnostic commands are for use by NetApp personnel only. Do you want to continue? {y|n}: y

Thanks mark

Weber, Mark A

28 Apr 28 Apr

1:32 a.m.

Thanks! That was the information I needed

From: Parisi, Justin [mailto:Justin.Parisi@netapp.com] Sent: Thursday, April 23, 2015 8:07 PM To: Weber, Mark A; Parisi, Justin; toasters@teaparty.net Subject: Re: cDOT cluster kernel-service

One of the clusters might have had it enabled as a holdover from a previous upgrade, or perhaps you had SAN enabled at one time.

They are both 8.2.3P1 So that failover is for SAN only? Neither cluster has FC or iSCSI enabled

thanks mark

That is for failover due to KCS being OOQ. Basically, if a node goes out of SAN quorum, it reboots.

What releases are your two clusters on? May just be a change in the default behaviors.

Anyone have a 4-node 2240 cluster running 8.2.x?

LCNA2::> set d

Warning: These diagnostic commands are for use by NetApp personnel only. Do you want to continue? {y|n}: y

Thanks mark

Tony Bar

1:54 a.m.

Mark -

Out of curiosity, do you have iSCSI/FCP licensed, even if not enabled or in use? Is this by any chance a 2240/255x based cluster? I'm curious about this & going to check on our 4 node 3220 cluster tomorrow to see if it's doing this too - we only use NFS so it'd be interesting to see if this is a default behavior.

Anthony Bar 650.207.5368tel:650.207.5368 | tbar@berkcom.commailto:tbar@berkcom.com

On Apr 27, 2015, at 6:31 PM, Weber, Mark A <mark-a-weber@uiowa.edumailto:mark-a-weber@uiowa.edu> wrote:

Thanks! That was the information I needed

From: Parisi, Justin [mailto:Justin.Parisi@netapp.com] Sent: Thursday, April 23, 2015 8:07 PM To: Weber, Mark A; Parisi, Justin; toasters@teaparty.netmailto:toasters@teaparty.net Subject: Re: cDOT cluster kernel-service

Right. So if only NAS, disable it. It’s only for SAN kernel quorum and automatic failover when OOQ. Storage failover (SFO) is different and is viewed via "storage failover show.” You do want SFO enabled.

One of the clusters might have had it enabled as a holdover from a previous upgrade, or perhaps you had SAN enabled at one time.

They are both 8.2.3P1 So that failover is for SAN only? Neither cluster has FC or iSCSI enabled

thanks mark

That is for failover due to KCS being OOQ. Basically, if a node goes out of SAN quorum, it reboots.

What releases are your two clusters on? May just be a change in the default behaviors.

Anyone have a 4-node 2240 cluster running 8.2.x?

If so, could you tell me if failover says enabled or disabled when you do a “cluster kernel-service show” in diag mode. I’ve had a performance case open for months on this cluster and I noticed recently that this cluster says failover is disabled but our 4 node 3220 cluster says enabled.

LCNA2::> set d

Warning: These diagnostic commands are for use by NetApp personnel only. Do you want to continue? {y|n}: y

Thanks mark _______________________________________________ Toasters mailing list Toasters@teaparty.netmailto:Toasters@teaparty.net http://www.teaparty.net/mailman/listinfo/toasters

Weber, Mark A

3:02 p.m.

Neither cluster has iSCSI or FCP licenses installed. The 2240 cluster was built by me (they were 7-mode systems) so I'm fairly certain it never had the licenses added. The 3220 cluster came with cDOT on it, so it is possible that it had demo licenses at some point. I tried to find my install log for the first two nodes, but it isn't where it should be so I haven't been able to check. The initial report from PS's NetAppDocs script lists only CIFS/NFS/SNAP[MIRROR|RESTORE|VAULT] installed.

From: Tony Bar [mailto:tbar@BERKCOM.com] Sent: Monday, April 27, 2015 8:54 PM To: Weber, Mark A Cc: Parisi, Justin; toasters@teaparty.net Subject: Re: cDOT cluster kernel-service

Mark -

Anthony Bar 650.207.5368tel:650.207.5368 | tbar@berkcom.commailto:tbar@berkcom.com

On Apr 27, 2015, at 6:31 PM, Weber, Mark A <mark-a-weber@uiowa.edumailto:mark-a-weber@uiowa.edu> wrote: Thanks! That was the information I needed

One of the clusters might have had it enabled as a holdover from a previous upgrade, or perhaps you had SAN enabled at one time.

They are both 8.2.3P1 So that failover is for SAN only? Neither cluster has FC or iSCSI enabled

thanks mark

That is for failover due to KCS being OOQ. Basically, if a node goes out of SAN quorum, it reboots.

What releases are your two clusters on? May just be a change in the default behaviors.

Anyone have a 4-node 2240 cluster running 8.2.x?

LCNA2::> set d

Warning: These diagnostic commands are for use by NetApp personnel only. Do you want to continue? {y|n}: y

Thanks mark _______________________________________________ Toasters mailing list Toasters@teaparty.netmailto:Toasters@teaparty.net http://www.teaparty.net/mailman/listinfo/toasters

3531

Age (days ago)

3536

Last active (days ago)

toasters@lists.teaparty.net

6 comments

3 participants

tags (0)

participants (3)

Parisi, Justin
Tony Bar
Weber, Mark A