Marcel, Clifton - thanks for the replies.
Marcel - I’ll send you that information pretty soon.
For some more information (I’ll try not to make it long) I’ll say the following:
1> I did try the description but that had no effect either. On the server that worked, it reversed all the changes on the switch. I’ve noticed that hp switches (the older ones anyway) take advantage of the description while the newer ones take advantage of the alias, so sometimes I have to make host-rules and not folder rules in those instances.
2> I use this option for auto-inventorying for varying reasons (check out the settings in the link) - I don’t use it everywhere, but I’m piloting it out at 2 sites for testing: http://d.pr/i/4DJh
So, after Clifton sends the email last night I look at the All Hosts section and it looks like for whatever reason, the settings are finally looking like they are starting to update because I see this: http://d.pr/i/QgkY
Ok, cool - on the working server, I’ll see that happen and usually during the next scheduled check, the services correct themselves since I have it set to fix them and apply the changes. Well, it hasn’t happened on this server yet and it’s ran through a longer-than-8-hour-cycle where the changes should have automatically applied themselves.
3> I have gone through the
Bulk inventory Add unmonitored & remove vanished services+Include subfolders+Do not use cached data+Do FULL SNMP scan
Then even all the same options except subbing the first option with Tabula Rasa.
It completes successfully but I still have a lot of UNKNOWNs when I look at the All hosts page - they haven’t changed at all.
4> Go to all hosts page, and hit an action menu on one of the switches and reschedule the check - nothing corrects itself.
5> OPEN the host and hit the action item on Check_MK inventory and the UNKNOWN goes away and the host is corrected. Weird still is that when I open the host, it shows all of the services are correct, even when the UNKNOWN status is present. I see the alias names properly. It states there are vanished services, but when you do a full scan on the device, it finds nothing - all of the information is completely accurate…
6> There aren’t any stale services; it almost seems like something is stuck or not refreshing. Rebooted the monitoring server as well, but to no avail. Seems like it’s churning through the process, but slowly - like something is stuck or not applying immediately.
I’m running 1.2.8p15 raw
···
–
Brian Binder
On January 12, 2017 at 12:04:44 AM, Marcel Schulte (schulte.marcel@gmail.com) wrote:
Hi Brian,
Could you please send a snmpwalk file for each if the
both switches? That’s the resulting file from “cmk -v --snmpwalk
{hostname}”.
If that’s not possible, can you send it as direct
mail, without mailing list?
And what CMK version are you using?
Regards,
Marcel
Brian Binder brian.binder@gmail.com
schrieb am Do., 12. Jan. 2017 03:05:
Anyone ever experience this not working for
them? I can’t place what I’m missing. I have 2 servers
and the rules are the same for this area…yet one of them will not
display the alias name for the service and I don’t know what the
issue is. Has this happened to you guys at all?
1.2.8p15 on both servers
Working rule on both look like this, but only one
of them displays the service correctly: http://d.pr/i/HBtH
So on the working server, I get
this: http://d.pr/i/lAsx
On the non-working one, I get this: http://d.pr/i/zaRT
I have removed the autochecks, I have
reinventoried the switches with cmk- ivp, -Ivp and -IIvp and then
-R
No matter what I do, the rule seems like it never
hits. I was hoping it would shot in the column on the active
check parameters column, but it doesn’t seem to show that on either
server.
I’m pretty interested in why it’s not
working. When I can use the alias as the service, I have a
nice way to only inventory ports that have been named outside of
their default names, which was something I posted on inquiry for
awhile back. Now when a port is being used for something
other than a workstation, it picks up the name on inventory since
it’s not the default interface name.
Thanks for any ideas,
–
Brian Binder
checkmk-en mailing list
checkmk-en@lists.mathias-kettner.de
http://lists.mathias-kettner.de/mailman/listinfo/checkmk-en