Forum Discussion
Terje_Gravvold
Nimbostratus
NimbostratusGraphing BigIP v9.x SNMP statistics with Cacti
In conjunction with the latter posts on SNMP mibs and SNMP monitoring on BigIP I want to do my small contribution. A while ago I spent some late hours to dig into the SNMP mibs of BigIP, my goal was to make graphs of BigIP internals in Cacti.
What's Cacti you wonder (www.cacti.net):
acti is a complete network graphing solution designed to harness the power of RRDTool's data storage and graphing functionality. Cacti provides a fast poller, advanced graph templating, multiple data acquisition methods, and user management features out of the box. All of this is wrapped in an intuitive, easy to use interface that makes sense for LAN-sized installations up to complex networks with hundreds of devices.
In other words, just the ting I need to make graphs on a budget! So what I've made is a set of templates that could easily be imported into Cacti for BigIP graphing fun...
Heres that the template currently provides:
- F5 BigIP - Chassis Temperature
- F5 BigIP - Global Connections
- F5 BigIP - Global HTTP 2xx/3xx Responses
- F5 BigIP - Global HTTP Requests
- F5 BigIP - Global HTTP Response Errors
- F5 BigIP - Global HTTP Responses
- F5 BigIP - Global packets/sec
- F5 BigIP - Interface Errors/Drops/Collisions
- F5 BigIP - Interface Traffic (bits/sec)
- F5 BigIP - SSL TPS
- F5 BigIP - TMM memory
- F5 BigIP - VS Connection Rate
- F5 BigIP - VS Connections
- F5 BigIP - VS Traffic (bits/sec)
And here is the link for those of you that got curious:
Click here
So why do I bother you with this SPAM?
1. I want to do my duty and contribute something back to F5 who has "given" me this fantastic product...
2. I want to see if I could get some smart brains on this forum to contribute additions to the template to get it complete. I'm, a) to lasy to do the work my selves, b) Don't have the time, your choise. Either way. take a look at Cacti and this template if you have the time...
Hint: iRules in combination with this template seems temptating to me :-)
Best regards
Terje Gravvold
33 Replies
HamishCirrocumulus
Posted By akiwerski on 01/05/2009 12:07 PM
Is it possible to get at the statistics for the individual TMM's on CMP platforms, such as the 8800, via SNMP/Cacti? I'd like to not only graph the CPU/TMM usage (which I currently do now), but also graph the individual TMM instances, like the performance graphs in the LTM GUI show.
/Alex K.
You might be in luck. There's a new table in the mibs downloaded from a 9.4.5 LTM box. Called sysTmmStat, that looks like it includes stats for individual tmm instances (indexed by tmmID, it includes the PID etc).
I don't have a box capable of running multiple tmm's though to test it...
H
chandru_14815Thanks Hamish!!
Can you help me how to acheive this?
ingardDoes anyone know how I can graph 2,3,4,5xx per pool or vserver? Is it possible at all?- HamishYep. The templates that were first uploaded around here somewhere will do that... They're on the first page of this thread
- ingardhm, maybe im misunderstanding something but it seems there are only global stats, and not per pool ?
- JRahm
Admin
Yeah, that's not going to work with the oids in that template, these are profile-based only and gather for all virtuals with that profile. However, you could build a stats profile and use an Irule to populate them. These oids are in the F5-BIGIP-LOCAL-MIB.txt beginning with ltmUserStatProfile. - JRahmSure, Joe wrote a tech tip a while back on this: Click here
- Hamish
Posted By citizen_elah on 09/11/2009 7:37 AM
Yeah, that's not going to work with the oids in that template, these are profile-based only and gather for all virtuals with that profile. However, you could build a stats profile and use an Irule to populate them. These oids are in the F5-BIGIP-LOCAL-MIB.txt beginning with ltmUserStatProfile.
Sorry. I pointed at the wrong template. I did one to pull HTTP stats from a profile too. I'll have a look and see if I can find it. (I use it for compression stats mostly).
H 
DaveC_21078Altostratus
This is good info. I'm trying to graph VS current connections, but the OIDs are all returning zero. OIDs like vs EphemeralCurrConns *
and vs_addr ClientCurrConns *. snmpwalk returns all my virtual servers correctly, but the counters are all at zero. I can get total connections fine. Does something need to be turned on to be able to retieve current connection numbers? I tried both virtual servers and pools with the same results. Thanks- ingardI just had a case with support about retrieving stat profile counters via snmp. It turns out there is a bug which they fixed with a HotFix for me.
