Hello all.
We have recently noticed that machine tracker will report duplicate arps on every entry on networks running hsrp. Looking at the gwportprefix.virual, it is set to 't' for the 'virtual' interfaces and 'f' for the 'real' interfaces. Seems right. I believe only the active interface should be collected, or have I missed something? Any magic wands or switches? The 2 'real' interfaces are entered in the seed database, not the virtual. We do not use the NTNU port description convention. We are running nav 4.6.1 and the last schema update is 4.6.56.
Any thoughts here?
Regards
Hello all.
We have recently noticed that machine tracker will report duplicate arps on every entry on networks running hsrp. Looking at the gwportprefix.virual, it is set to 't' for the 'virtual' interfaces and 'f' for the 'real' interfaces. Seems right. I believe only the active interface should be collected, or have I missed something? Any magic wands or switches? The 2 'real' interfaces are entered in the seed database, not the virtual. We do not use the NTNU port description convention. We are running nav 4.6.1 and the last schema update is 4.6.56.
Any thoughts here?
Regards
A little note. We had only the loopback interface in seed database first, with the same duplicates.
On Tue, 21 Feb 2017 12:33:08 +0000 Jan Sigurd Refvik j.s.refvik@usit.uio.no wrote:
We have recently noticed that machine tracker will report duplicate arps on every entry on networks running hsrp.
Hi Jan Sigurd,
NAV will happily dump ARP and ND cache tables from any router you are monitoring, regardless of whether it works in tandem with other routers on the same subnet. Interface config doesn't enter into it.
The source router is part of NAV's arp record, so the key for identifying a unique, open record is basically router+IP+MAC. If the source router is different, there will be a "duplicate" record for the same IP+MAC combo.
I'm not aware of whether the various router redundancy protocols guarantee that all cooperating routers will have the same exact set of ARP/ND records available at any given time.
Hi Morten.
Thanks for your fast reply. I do agree with you on this, but on the other hand it seems overkill as Nav knows about the virtual and real interfaces so these can be considered before dumping the arp tables. This might end up in a feature request. ;-)
mvh
-----Original Message----- From: Morten Brekkevold [mailto:morten.brekkevold@uninett.no] Sent: Tuesday, February 21, 2017 4:40 PM To: Jan Sigurd Refvik j.s.refvik@usit.uio.no Cc: nav-users@uninett.no Subject: Re: Duplicate arps on prefixes with hsrp.
On Tue, 21 Feb 2017 12:33:08 +0000 Jan Sigurd Refvik j.s.refvik@usit.uio.no wrote:
We have recently noticed that machine tracker will report duplicate arps on every entry on networks running hsrp.
Hi Jan Sigurd,
NAV will happily dump ARP and ND cache tables from any router you are monitoring, regardless of whether it works in tandem with other routers on the same subnet. Interface config doesn't enter into it.
The source router is part of NAV's arp record, so the key for identifying a unique, open record is basically router+IP+MAC. If the source router is different, there will be a "duplicate" record for the same IP+MAC combo.
I'm not aware of whether the various router redundancy protocols guarantee that all cooperating routers will have the same exact set of ARP/ND records available at any given time.
-- mvh Morten Brekkevold UNINETT
-
Jan Sigurd Refvik -
Morten Brekkevold