2009/6/15 Robert Story <[email protected]>:
> if internal to the cache helper,  that would mean OIDs in the conf file, since
> that's all the cache helper is passed..

No - it means that MIB object names in the config file would be converted
to numeric OIDs, presumably by the config parsing routine.

Internally the comparisons would be of numeric OIDs, yes.
But the user visible config could use names or numbers.


> DS> That way *all* cached tables would pick up run-time configuration
> DS> automatically, without needing to update per-module code.
>
> I don't know if that's possible. Some modification will be needed... For
> example, the ifTable creates the cache in once place with a generic default
> time, but then sets the final value in another place. An idiosyncrasy of it
> being MfD code..

No Comment :-)

(But I do think this is a less than ideal approach - specifying the timeout
in one place would be clearer, IMO)



> DS> Of course, this should also handle the persistency of SETing
> DS> these cache timeouts.
>
> persistence is a whole other wrinkle. If the default is 30, the conf file sets
> it to 15, but at runtime it is set to 45, and the agent shuts down. When the
> agent starts again, what value should it use?

   15

With the SET request to change the configured value having being rejected
as not-writable.   (c.f. sysLocation, etc)

> I guess rows with values from
> the conf file could be make read-only...

Exactly.

> But persistence could be done later, don't you think?

Yes - that might be sensible.
   (Though this should probably be straightforward anyway)

Dave

------------------------------------------------------------------------------
Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensing option that enables unlimited
royalty-free distribution of the report engine for externally facing 
server and web deployment.
http://p.sf.net/sfu/businessobjects
_______________________________________________
Net-snmp-coders mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/net-snmp-coders

Reply via email to