cluther wrote:
> It sounds like the raw value coming back through SNMP might not be
> hex, but snmpwalk is trying to be helpful and displaying it this way.
> If this is the case you should just make sure that you setup your data
> point as a COUNTER type. Then your graph will show you bytes/sec.
>
> Make sure you delete the corrosponding RRD file in /opt/zenoss/perf/
> Devices/deviceName/ after you change the type of the data point so
> that it can be recreated with the proper type.
I did try that just now, but no luck. The graph just sits there reading 'nan'
after several hours of collection. This is from the actual MIB -
proxyStatsObjCacheDataXfered OBJECT-TYPE
SYNTAX OCTET STRING (SIZE (8))
ACCESS read-only
STATUS mandatory
DESCRIPTION
"Total amount of object data transfered to browsers in Kbytes."
::= { proxyStatsObjCache 6 }
So as best I can tell with my limited experience (a days worth) in reading
MIBs, it really is a hex string. Most other values are SYNTAX Counter, the
only reason I can imagine this one is not is just because it gets stupid huge -
I've seen the server happily get to more than a few terabytes transferred, and
expressed in Kbytes, that overruns a 32bit integer, right? I think. Maybe.
It certainly would eventually, assuming the server was perfectly stable.
To jmp242 - I will take a look at command datasource + snmpwalk, I'm just a bit
lost how to do that actual conversion, but what the hey, it certainly sounds
like some fun fumbling around to look forward to.
Thanks guys! If anyone happens to be a script wizard that could make it look
real easy, though, I wouldn't say 'no'. =)
-------------------- m2f --------------------
Read this topic online here:
http://community.zenoss.com/forums/viewtopic.php?p=25140#25140
-------------------- m2f --------------------
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
