Unsolved
1 Rookie
•
14 Posts
0
18
SNMP Authentication Error Traps
I keep getting these alerts in OME. Most things I've read say it is due to an incorrect community name, but the name is just fine. The only resolution I can find is to uncheck the Send SNMP authentication trap in the Windows SNMP service. I'd rather not have to do that for a whole bunch of servers if possible.
This only happens periodically. It can go for a few weeks without any of these traps, then I get a bunch. I still run Open Manage Essentials in parallel with OME 4.1, and Essentials does not have this problem.
Just wondered if anyone has any better ideas before I go and modify the SNMP properties on every server.
DELL-Erman O
Moderator
Moderator
•
2.2K Posts
1
May 14th, 2024 09:29
Hello,
Make sure that the SNMP community string set up in OME is an exact match with the one on the servers. Even a small mistake like a typo could lead to an authentication error. Also, check that OME is using the same SNMP version—be it v1, v2c, or v3—as the one set on the servers. If they don’t match, it might cause authentication problems.
Now, even though Open Manage Essentials isn’t setting off these traps, it’s possible that a different SNMP management tool is trying to access the servers with the wrong credentials when these alerts pop up. This could be misleading OME.
It’s also a good idea to comb through the OME logs for any detailed error messages that occur at the same time as the authentication failures. These messages can often give you more specific hints about what’s going wrong.
If you’re still running into issues and looking for a quick fix, you have the option to turn off SNMP authentication traps on the particular servers that are giving you trouble. This won’t stop the alerts completely, but it will keep them quiet while still letting OME manage those servers.
Lastly, take a look at how often OME is polling SNMP. If you increase the interval between polls, it might prevent the system from getting its wires crossed with other tools, assuming that’s what’s causing the issue.
Hope that helps!
Nobodyimportant
1 Rookie
1 Rookie
•
14 Posts
0
May 24th, 2024 22:18
@DELL-Erman O Thanks for your reply. I didn't get an email so I didn't know anyone replied. Probably due to my profile settings.
I think my discovery settings might be part of the issue. Some devices are listed by iDrac IP and some by server IP. The SNMP errors are from devices OME regards as unknown because they come from the component OME hasn't discovered for some reason, meaning either the server or the iDrac.
I prefer to discover/manage things by server/DSM agent, but it seems 4.x wants to push you more to the iDrac. I guess that's because most people virtualize, but we are 99% physical servers. It doesn't work well for us. I also find the iSM is more hassle than it is worth.
I have been selecting WSMan/Redfish and SNMP as my discovery protocols, and using the whole IP range for each subnet, e.g 172.17.141.0/24. I figured that would catch iDracs and servers, but it seems to be only partially successful. I tried SSH on a few servers and it works well, but I don't want to set up SSH on every server, and some are too old anyway.
I'm thinking maybe I need to trash my whole installation and start again. I don't think I will be able to get OME to work the way I want it to though.