I think it's better to leave them as an interface with multiple IP
addresses, from a monitoring/management point of view, as they are not
really separate physical interfaces.
Perhaps it could be implemented as an option to allow folks to switch
between whichever suits them best?
-- Rick Hodger Bytel Ltd. W: http://www.bytel.net.uk/ T: +44 (0)28 9045 6111 F: +44 (0)28 9045 6123 > -----Original Message----- > From: Victor Kirhenshtein [mailto:victor_at_opticom.lv] > Sent: 02 September 2005 15:26 > To: NetXMS Users > Subject: Re: [netxms-users] linux subagent - Net.InterfaceList > > > I've just checked how agent can handle aliases created with ip command: > > monitor:~# ip addr add 10.0.7.1/16 dev eth0 > monitor:~# ip addr list > 1: lo: <LOOPBACK,UP> mtu 16436 qdisc noqueue > link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 > inet 127.0.0.1/8 brd 127.255.255.255 scope host lo > 2: eth0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 100 > link/ether 00:07:e9:60:1c:5f brd ff:ff:ff:ff:ff:ff > inet 10.0.0.4/16 brd 10.0.255.255 scope global eth0 > inet 10.0.7.1/16 scope global secondary eth0 > monitor:~# nxget 127.0.0.1 -l Net.InterfaceList > 1 127.0.0.1/8 1 000000000000 lo > 2 10.0.0.4/16 1 0007E9601C5F eth0 > 2 10.0.0.4/16 1 0007E9601C5F eth0 > > So agent can see that threre is another IP address on interface, but > cannot detect it correctly. Test program can detect all aliases > correctly. > > And I have one topic to discuss based on that aliases problem: > currently, NetXMS handles multiple addresses on an interface by creating > multiple interface objects related to node. Should we leave this > behavior as it is, or more logical to have one interface object with > multiple IP addresses? > > Best regards, > Victor > > > -----Original Message----- > > From: Christoffer Blindheim [mailto:netxms_at_christoffer.totalnett.no] > > Sent: Friday, September 02, 2005 5:03 PM > > To: NetXMS Users > > Subject: Re: [netxms-users] linux subagent - Net.InterfaceList > > > > > > > > Never used the ip command to setup aliases... don't now if it > > works with > > Net.InterfaceList. But this bug I think were talking about virtual > > interfaces. See my attatched post. > > > > Christoffer Blindheim wrote: > > > > >> > > >> nxget -a plain -s mysecret -l servername Net.InterfaceList gives: > > >> 500: Internal error > > >> > > >> Tried this on two diffrent gentoo servers, but no luck. The only > > >> gentoo server that gives me a interface list is the > > managementserver, > > >> but this has the alias problem... > > >> > > >> Alias: > > >> 1 127.0.0.1/8 1 000000000000 lo > > >> 2 xxx.yyy.zzz.ccc/27 1 000EA62D798E eth0 > > >> 3 10.20.1.19/16 1 0002B39A806B eth1 > > >> > > >> my ifconfig |grep eth > > >> eth0 Link encap:Ethernet HWaddr 00:0E:A6:2D:79:8E > > >> eth0:1 Link encap:Ethernet HWaddr 00:0E:A6:2D:79:8E > > >> eth0:2 Link encap:Ethernet HWaddr 00:0E:A6:2D:79:8E > > >> eth0:3 Link encap:Ethernet HWaddr 00:0E:A6:2D:79:8E > > >> eth0:4 Link encap:Ethernet HWaddr 00:0E:A6:2D:79:8E > > >> eth1 Link encap:Ethernet HWaddr 00:02:B3:9A:80:6B > > >> eth1:1 Link encap:Ethernet HWaddr 00:02:B3:9A:80:6B > > >> > > >> However this system is running 2.6 kernel, could this cause it? My > > >> 2.4 rh-7.2 is responding correctly. > > >> > > >> Do you have a small program that utilizes the if_nameindex > > function > > >> that I could try to run on my systems for debugging? > > >> > > >> thanks. > > > > > > > Rick Hodger wrote: > > > > >I think what he's talking about is aliases, and not virtual > > interfaces > > >(ala eth0:1, eth0:2, eth0:3 etc). Aliases use the "ip" > > command and do > > >not show up in the "ifconfig" output. > > > > > >Same result, just a different method. > > > > > > > > > >Received on Fri Sep 02 2005 - 17:32:24 EEST
This archive was generated by hypermail 2.2.0 : Fri Sep 02 2005 - 17:41:35 EEST