SNMP host down status but can snmpwalk it

[pennmanor]

I have an MGE Pulsar 2200 with a snmp card installed. I am able to do an snmpwalk on it from the server that cacti is running on

[root@it etc]# snmpwalk -v 1 -c public -O n 172.19.186.13
.1.3.6.1.2.1.1.1.0 = STRING: Pulsar EX 2200 RT
.1.3.6.1.2.1.1.2.0 = OID: .1.3.6.1.4.1.705.1.2
.1.3.6.1.2.1.1.3.0 = Timeticks: (1512234497) 175 days, 0:39:04.97

If I go in the device in cacti I see at the top that it has pulled some info

HS MDF UPS (172.19.186.13)
SNMP Information
System: Pulsar EX 2200 RT
Uptime: 1512296025 (175 days, 0 hours, 49 minutes)
Hostname: hsupsmdf01
Location: hs mdf left rack
Contact: Computer Room Manager

but the status is listed as down and I do get an snmp poller error in the log

01/26/2007 01:30:03 PM - CMDPHP: Poller[0] DEBUG: SQL Exec: "update host set status = '1', status_event_count = '313', status_fail_date = '0000-00-00 00:00:00', status_rec_date = '0000-00-00 00:00:00', status_last_error = 'Host did not respond to SNMP', min_time = '9.99999', max_time = '0.00000', cur_time = '0.00000', avg_time = '0.00000', total_polls = '314', failed_polls = '314', availability = '0' where hostname = '172.19.186.13'"
01/26/2007 01:30:03 PM - CMDPHP: Poller[0] Host[34] SNMP: Host did not respond to SNMP

any idea where I should look now ? I am running just the snmp poller and not the ping poller but it didn't work when both were enabled

[gandalf]

Please try my NaN Debugging Stuff
Reinhard

[pennmanor]

Been there and done that. All of my other hosts work fine but none of my MGE units work.

The thing that I find strange is that the uptime stats on the top of the device info page is updating so that must mean that it is communicating

System: Evolution 1500
Uptime: 260703487 (30 days, 4 hours, 10 minutes)
Hostname: mmsupsmdf01
Location: rack in tech office
Contact: Computer Room Manager


SNMP Information
System: Evolution 1500
Uptime: 260710690 (30 days, 4 hours, 11 minutes)
Hostname: mmsupsmdf01
Location: rack in tech office
Contact: Computer Room Manager

[pennmanor]

Ok, I think I may be onto something. I found in this thread that the snmp poller checks for .1 oid .

http://forums.cacti.net/viewtopic.php?t=19267

Strange thing is that if I snmpwalk one of the mge units for just .1 I get

[root@it var]# snmpwalk -v 1 -c public -O n 172.19.186.13 .1
Timeout: No Response from 172.19.186.13


but if I snmpwalk .1.3 I get data back. I walked other devices for .1 and they worked. So, I guess my question is now how do I change the poller to look for .1.3 instead of just .1 ?

[pennmanor]

I found it , line 284 of lib/ping.php . I changed to .1.3 and ran the poller again. All is good now.

[gandalf]

Fine, then. But it should have been an snmpgetnext to .1 and not an snmpwalk. But I suppose this would yield the same results ...
Reinhard

[pennmanor]

I got the same result with snmpgetnext.

[root@it lib]# snmpgetnext -c public -v 1 172.19.186.13 .1.3
SNMPv2-MIB::sysDescr.0 = STRING: Pulsar EX 2200 RT
[root@it lib]# snmpgetnext -c public -v 1 172.19.186.13 .1
Timeout: No Response from 172.19.186.13.
[root@it lib]#

[gandalf]

Oh yes, I expected that. Perhaps I did not make myself plain enough, as english is not my native language. I only wanted to correct tyour statement above, as cacti performs an snmpgetnext with latest patches. And I expected that this will fail as well.
Reinhard

[hipsterdoofus]

I had the same problem with an MGE ups...changing the number in the ping.php fixed it...thanks!

[TheWitness]

There is an official patch for this on the web site.

TheWitness