On Tue, 07 Sep 2010 15:08:26 -0700 Wes wrote: WH> >>>>> On Tue, 7 Sep 2010 11:22:08 -0700, Stephen Hemminger <shemmin...@vyatta.com> said: WH> WH> SH> I have fix to make it keep overrides. But I did not implement WH> SH> setting kernel ifalias because on most systems snmpd runs as WH> SH> non-root (no cap_netadmin) and is therefore not allowed to set the WH> SH> value in sysfs (or via netlink).
I'm not sure that snmpd runs as non-root on most systems. Regardless, I think we should provide the ability to set it, and let the system admin decide if they want to run as root or set cap_netadmin. WH> Then we should probably discuss if you SET the ifAlias then snmpd simply WH> reports that value back in the future, regardless of what the kernel WH> might report through other ways. EG an alias would be: WH> WH> 1) the last value set for that interface WH> 2) the super-cool string created from kernel info (from the patch) if possible WH> 3) an empty string I'm not a big fan of this idea. If we pull data from the kernel, then we should keep it in sync. Keeping our own value that doesn't match feels wrong. WH> [in fact, it's probably questionable as to whether the patch is legal in WH> the first place since it says it should be a zero-length string until WH> set] Although it might violate the letter of the law, I think it is in keeping with the spirit of the law, which is simply to provide a long term alias. And if the kernel does the dirty work of keeping track of the alias even if the ifIndex changes (which we don't currently do), all the better. ------------------------------------------------------------------------------ Start uncovering the many advantages of virtual appliances and start using them to simplify application deployment and accelerate your shift to cloud computing http://p.sf.net/sfu/novell-sfdev2dev _______________________________________________ Net-snmp-coders mailing list Net-snmp-coders@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/net-snmp-coders
