dbTalk Databases Forums  

2 Node Active/Passive - Failover when switch rebooted.

microsoft.public.sqlserver.clustering microsoft.public.sqlserver.clustering


Discuss 2 Node Active/Passive - Failover when switch rebooted. in the microsoft.public.sqlserver.clustering forum.



Reply
 
Thread Tools Display Modes
  #1  
Old   
David Morgan
 
Posts: n/a

Default 2 Node Active/Passive - Failover when switch rebooted. - 11-07-2007 , 02:08 PM






Hi

Despite being directly connected with a crossover cable and each server
being connected in to the switch, they seem to think that they need to
failover when the switch goes off line. Why?

Surely they should only be checking connectivity if the heartbeat fails over
the direct connection. Then, they attempt to use the LAN connection.

Have I missed something here?

Thanks

David



Reply With Quote
  #2  
Old   
Geoff N. Hiten
 
Posts: n/a

Default Re: 2 Node Active/Passive - Failover when switch rebooted. - 11-07-2007 , 02:21 PM






The Physical port is dependent on the switch port.
The IP address is dependent on the Physical port
The Network Name is dependent on the IP address (Cluster Manager)
The Service is dependent on the Network Name (Cluster Manager).

The switch goes offline and starts a cascade of failure. This is exactly
what is expected.

--
Geoff N. Hiten
Senior SQL Infrastructure Consultant
Microsoft SQL Server MVP


"David Morgan" <microsoft_newsgroups (AT) NOdavidmorgan (DOT) SPAMme.uk> wrote in
message news:u6$pDoXIIHA.6068 (AT) TK2MSFTNGP05 (DOT) phx.gbl...
Quote:
Hi

Despite being directly connected with a crossover cable and each server
being connected in to the switch, they seem to think that they need to
failover when the switch goes off line. Why?

Surely they should only be checking connectivity if the heartbeat fails
over the direct connection. Then, they attempt to use the LAN connection.

Have I missed something here?

Thanks

David



Reply With Quote
  #3  
Old   
Tom Moreau
 
Posts: n/a

Default Re: 2 Node Active/Passive - Failover when switch rebooted. - 11-07-2007 , 04:38 PM



Consider this. The cluster has to be able to determine if it is servicing
the public. It doesn't do that through the crossover. Rather, it has to
try to communicate via the public (non-heartbeat) NIC(s). If the switch
goes offline, that tells the node that owns the IP that it can't communicate
with the outside world. Hence, the failover.

Conversely, if you cut the crossover cable, there should not be a failover,
since the cluster can communicate via the public NIC(s), if so configured.

--
Tom

----------------------------------------------------
Thomas A. Moreau, BSc, PhD, MCSE, MCDBA, MCITP, MCTS
SQL Server MVP
Toronto, ON Canada
https://mvp.support.microsoft.com/profile/Tom.Moreau


"David Morgan" <microsoft_newsgroups (AT) NOdavidmorgan (DOT) SPAMme.uk> wrote in
message news:u6$pDoXIIHA.6068 (AT) TK2MSFTNGP05 (DOT) phx.gbl...
Hi

Despite being directly connected with a crossover cable and each server
being connected in to the switch, they seem to think that they need to
failover when the switch goes off line. Why?

Surely they should only be checking connectivity if the heartbeat fails over
the direct connection. Then, they attempt to use the LAN connection.

Have I missed something here?

Thanks

David



Reply With Quote
Reply




Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off



Powered by vBulletin Version 3.5.3
Copyright ©2000 - 2012, Jelsoft Enterprises Ltd.