Spine using snmp v2c constantly returns "value: U"

[Muffinman]

Hi,

I am using Spine in version 0.8.7e and snmp v2. On many hosts I get a "value: U" instead of a real number. Like this:

11/13/2009 03:36:58 PM - SPINE: Poller[0] Host[346] DS[8057] SNMP: v2: myhost.mydomain.tld, dsname: traffic_out, oid: .1.3.6.1.2.1.2.2.1.16.4, value: 1469757011
11/13/2009 03:36:58 PM - SPINE: Poller[0] Host[346] DS[8057] SNMP: v2: myhost.mydomain.tld, dsname: traffic_in, oid: .1.3.6.1.2.1.2.2.1.10.4, value: 318850776
11/13/2009 03:36:58 PM - SPINE: Poller[0] Host[346] DS[8064] SNMP: v2: myhost.mydomain.tld, dsname: cache_objects, oid: .1.3.6.1.4.1.3495.1.3.1.7, value: U
11/13/2009 03:36:58 PM - SPINE: Poller[0] Host[346] DS[8065] SNMP: v2: myhost.mydomain.tld, dsname: http_data_received, oid: .1.3.6.1.4.1.3495.1.3.2.1.4, value: U

And when I perform a snmpwalk, everything is fine:

snmpwalk -v2c -c public myhost.mydomain.tld -Cc .1.3.6.1.2.1.2.2.1.16.4 IF-MIB::ifOutOctets.4 = Counter32: 1866147654
snmpwalk -v2c -c public myhost.mydomain.tld -Cc .1.3.6.1.2.1.2.2.1.10.4 IF-MIB::ifInOctets.4 = Counter32: 678136459
snmpwalk -v2c -c public myhost.mydomain.tld:3401 -Cc .1.3.6.1.4.1.3495.1.3.1.7 SNMPv2-SMI::enterprises.3495.1.3.1.7.0 = Gauge32: 848284
snmpwalk -v2c -c public myhost.mydomain.tld:3401 -Cc .1.3.6.1.4.1.3495.1.3.1.4 SNMPv2-SMI::enterprises.3495.1.3.1.4.0 = INTEGER: 4703

As you can see there are values available. I did enter the Squid Port 3401 in cacti.

I also opened a bug report as there is such a topic for snmp v3. http://bugs.cacti.net/view.php?id=1572

Has anybody a workaround?

TIA

Muffinman

[gandalf]

I'm quite stumped.
Traffic should only be reported when querying against port 161 and squid data only on 3401. Did you enter the squid proxy directive?
R.

[Muffinman]

The standard port is still 161.
In the data template for the squid things I set the port to 3401.

(Of course I allowed on the squid host the squid to report to port 3401 to my cacti host.)

Did this answer your question?

Thanks!

[farkast21]

Hi!

The solution is to change the SNMP OIDs to end with 0, as you see it in the snmpwalk results.
I had a similar problem too, and this helped in my case.

[Muffinman]

I will check this when I am back from my holiday.
Thanks for this quick reply!

[gandalf]

In the data template for the squid things I set the port to 3401.

Please show us how you did this; a screenshot will do
R.

[TheWitness]

Gandolf, there is a bug here. We should discuss the solution. I encountered it when doing the Lossless Reindexing patch. It's in the bowels of the poller_item table generation.

TheWitness

[Muffinman]

[quote="gandalf a screenshot will do
[/quote]

Here you are:

[Muffinman]

Mea culpa!

I have two squids and one has SNMP compiled in and the other has not.
I asked the wrong one.
When I ask the right one the output is like this:

Host[180] DS[7067] SNMP: v2: myhost.mydomain.tld, dsname: traffic_out, oid: .1.3.6.1.2.1.2.2.1.16.2, value: 2840026622
Host[180] DS[7067] SNMP: v2: myhost.mydomain.tld, dsname: traffic_in, oid: .1.3.6.1.2.1.2.2.1.10.2, value: 3978745110
Host[180] DS[10257] SNMP: v2: myhost.mydomain.tld, dsname: serv_requests, oid: .1.3.6.1.4.1.3495.1.3.2.1.10.0, value: 1617871
Host[180] DS[10253] SNMP: v2: myhost.mydomain.tld, dsname: http_data_saved, oid: .1.3.6.1.4.1.3495.1.3.2.2.1.10.5, value: 0
Host[180] DS[10254] SNMP: v2: myhost.mydomain.tld, dsname: http_data_sent, oid: .1.3.6.1.4.1.3495.1.3.2.1.5, value: 24170340
Host[180] DS[10255] SNMP: v2: myhost.mydomain.tld, dsname: musage_size, oid: .1.3.6.1.4.1.3495.1.3.1.3.0, value: 89264

When I ask the one without SNMP, I get the "value: U" -- and that doesn't suprise me...

Mea maxima culpa!

I will close my bug report myself...