The section of the Bundled Agents Ref Guide that shows Failback has a note above it that these are optional for Base mode. You are using MPathd mode, so Failback setting is ignored. Your issue is with Mpathd, not MNICB
nfs10 in.mpathd[9033]: [ID 620804 daemon.error] Successfully failed back to NIC e1000g0 ________________________________ From: veritas-ha-boun...@mailman.eng.auburn.edu [mailto:veritas-ha-boun...@mailman.eng.auburn.edu] On Behalf Of Mark Schipper Sent: Monday, October 18, 2010 12:54 PM To: Christian Gerbrandt; veritas-ha@mailman.eng.auburn.edu Subject: Re: [Veritas-ha] MultiNICB attribute question Hmmm... That's what I would have expected, but my cluster is clearly set: Failback = 0 And I had failback happen 4 times (IP switched interfaces 8 times) while a switch port was flapping a few days ago. Would have been desirable for the IP's to stay on the secondary NIC. Here's proof of failback: Oct 14 20:44:43 nfs10 in.mpathd[9033]: [ID 594170 daemon.error] NIC failure detected on e1000g0 of group mnic_pcsnfs3 Oct 14 20:44:43 nfs10 in.mpathd[9033]: [ID 832587 daemon.error] Successfully failed over from NIC e1000g0 to NIC e1000g2 Oct 14 20:45:03 nfs10 in.mpathd[9033]: [ID 299542 daemon.error] NIC repair detected on e1000g0 of group mnic_pcsnfs3 Oct 14 20:45:03 nfs10 in.mpathd[9033]: [ID 620804 daemon.error] Successfully failed back to NIC e1000g0 And proof that my VCS config uses mpathd: MultiNICB mnic_pcsnfs3 ( UseMpathd = 1 Device @nfs09 = { e1000g0 = 0, e1000g2 = 1 } Device @nfs10 = { e1000g0 = 0, e1000g2 = 1 } Device @nfs11 = { e1000g0 = 0, e1000g2 = 1 } ) Does this mean "Failback = 0" was ignored or over-ridden? From: veritas-ha-boun...@mailman.eng.auburn.edu [mailto:veritas-ha-boun...@mailman.eng.auburn.edu] On Behalf Of Christian Gerbrandt Sent: Monday, October 18, 2010 8:56 AM To: Mark Schipper; veritas-ha@mailman.eng.auburn.edu Subject: Re: [Veritas-ha] MultiNICB attribute question It is the opposite way: Table 3-13 Optional attributes for Base mode Optional attribute Description Failback If the value of the attribute is 1, the virtual IP addresses are failed back to the original physical interface whenever possible. A value of 0 disables this behavior. Type and dimension: integer-scalar Default: 0 From: veritas-ha-boun...@mailman.eng.auburn.edu [mailto:veritas-ha-boun...@mailman.eng.auburn.edu] On Behalf Of Mark Schipper Sent: 18 October 2010 16:40 To: veritas-ha@mailman.eng.auburn.edu Subject: [Veritas-ha] MultiNICB attribute question Gang, The default behavior of MultiNICB is to switch back to the primary interface if/when link is restored. Documentation I have read confirms this. And since the "Failover" attribute defaults to "0": [r...@db09:/]# haattr -display MultiNICB | grep Failback Failback [integer/scalar] = 0 I would like to verify that setting: Failback = 1 will prevent an interface "failback" when the primary NIC link is restored. Thanks ........................................................................... Mark Schipper | UNIX Engineer +1 (408) 688-3196 office | +1 (831) 325-9882 mobile Syniverse Technologies | We make mobile work. mark.schip...@syniverse.com<mailto:email.addr...@syniverse.com> | www.syniverse.com<http://www.syniverse.com/> ...........................................................................
_______________________________________________ Veritas-ha maillist - Veritas-ha@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-ha
