|
From: Buchan M. <bg...@st...> - 2011-08-10 15:46:44
|
On Wednesday, 10 August 2011 16:46:40 White, Bruce wrote: > Buchan, > > > > Thanks very much for the help. I started this template by copying the > cisco-6509 template. I first used the model option to point to the 6509 > template, and after having issues I copied the template and adjusted the > specs file so devmon finds the new cisco-n7000 template on its own. > The diff for the specs file are: > [...] > [11-08-10@08:30:14] No SNMP data found for ifName on Nexus-DR Note that if you are running anything but a relatively current svn snapshot, this message can be misleading. See below. > > [11-08-10@08:30:14] No SNMP data found for ifOutPktsSec on Nexus-DR > > [11-08-10@08:30:14] No SNMP data found for ifOutCollisions on Nexus-DR > > [11-08-10@08:30:14] Missing repeater data for ifName for if_col msg > > > > Similar messages appear for the other tests requiring the ifName table > data. > > > > When I ran the snmpwalk against our 6509 I get the needed data but the > index starts at 1 and is incremented by 1, unlike what I am seeing in > the data for the n7000 switch where the index to the ifName table > appears to have no rhyme or reason. This is the only difference I see > in the snmpwalk output and I thought it was the cause of my issues. I have seen a lot of other instances where this is not an issue, but your other issues below are common. > Based on your earlier email I guess that is not that issue. > > > > What other information can I provide that will be helpful? > > > > Here are the messages I get when I run devmon with the --debug flag > against the device with the copied cisco-6509 template with my mods that > work for cpu and memory and adding an ndxtst which I set up to try to > get this ifName stuff to work. I have edited the output down to what > appears to be the relevant pieces: > > > > [11-08-10@08:44:37] ---Initilizing devmon... > > [11-08-10@08:44:37] Node 0 reporting to lndrmn01.na.fellowes.com > > [11-08-10@08:44:37] Running under process id: 13315 > > [11-08-10@08:44:37] Entering poll loop > > [11-08-10@08:44:37] DEBUG TEMPLATES: running read_templates() > > [11-08-10@08:44:37] DEBUG TEMPLATES: read cisco:n7000:if_load template > > [11-08-10@08:44:37] DEBUG TEMPLATES: read cisco:n7000:if_err template > > [11-08-10@08:44:37] DEBUG TEMPLATES: read cisco:n7000:serial template > > [11-08-10@08:44:37] DEBUG TEMPLATES: read cisco:n7000:ndxtst template > > [11-08-10@08:44:37] DEBUG TEMPLATES: read cisco:n7000:if_stat template > > [11-08-10@08:44:37] DEBUG TEMPLATES: read cisco:n7000:temp template > > [11-08-10@08:44:37] DEBUG TEMPLATES: read cisco:n7000:cpu template > > [11-08-10@08:44:37] DEBUG TEMPLATES: read cisco:n7000:if_col template > > [11-08-10@08:44:37] DEBUG TEMPLATES: read cisco:n7000:if_dsc template > > [11-08-10@08:44:37] DEBUG TEMPLATES: read cisco:n7000:power template > > [11-08-10@08:44:37] DEBUG TEMPLATES: read cisco:n7000:memory template > > [11-08-10@08:44:37] DEBUG TEMPLATES: read cisco:n7000:fans template > I would recommend you disable all of the tests above that are not working, except for the if_load, and once it is green, enable them one-by-one. > [11-08-10@08:44:37] DEBUG CFG: running read_hosts > > [11-08-10@08:44:37] DEBUG SNMP: running poll_devices() > > [11-08-10@08:44:38] DEBUG SNMP: Dethawing data for Nexus-DR > > [11-08-10@08:44:38] DEBUG SNMP: Failed queries 0 > > [11-08-10@08:44:38] DEBUG SNMP: Failed queries 0 > > [11-08-10@08:44:38] DEBUG SNMP: Adding ID => 1 OID > =>1.3.6.1.4.1.9.9.48.1.1.1.6.1 > > [11-08-10@08:44:38] DEBUG SNMP: Pooling 1 oids > > [11-08-10@08:44:38] DEBUG SNMP: Failed queries 0 > > [11-08-10@08:44:38] DEBUG SNMP: Failed queries 0 > > [11-08-10@08:44:38] DEBUG SNMP: Failed queries 0 > > [11-08-10@08:44:38] DEBUG SNMP: Dethawing data for sw3750-24_it_dr_50 > > [11-08-10@08:44:47] SNMP: Queried Nexus-DR: expected 38, received 11 So, you had 27 failures here. The way devmon prior to (as-yet-unreleased) 0.3.1 worked is that after 3 empty responses (e.g. if the device doesn't support the OID requested), it would give up. In current svn, it only gives up after 3 empty responses that had an error message. I guess I should really try and get a new release out ASAP. Regards, Buchan |