Server RelayNotifier http errors

(imported topic written by JonLandis)

I upgraded my dev environment to 9.0.787.0 a few weeks ago. Since then, I’ve been getting frequent RelayNotifier http errors. I find that when they occur, I need to cycle the root server service in order for the environment to be functional again. Has anyone else seen these?

RelayNotifier4 (6208) - Error running task RelayNotifier4: HTTP Error 35: SSL connect error: error:1408F10B:SSL routines:SSL3_GET_RECORD:wrong version number

RelayNotifier7 (8572) - Error running task RelayNotifier7: HTTP Error 28: Timeout was reached: Connection timed out after 10000 milliseconds

The notifier numbers vary, but the results are the same.

(imported comment written by ItsAvi)

Hi,

was it solved in any way?

when I enabled debug logging I see this:

RelayNotifier10 (3940) - Exiting POST
https://relay:52311/notify
(237 ms)

why is it trying to use HTTPS?

(imported comment written by MarlonRojas)

Friend anyone resolve this issue?

Has anyone found the solution for this issue ?

‘Connection timed out’ may have been more of an issue with certain 9.0/9.1 versions of IEM. There were a umber of users seeing the same issues and upgrading to 9.2 improved things. Anyone one to back that up? :sweat_smile:

In the meantime, you can try increasing the timeout by changing the setting “_BESData_Comm_TimeoutSeconds” to see if it helps, although this isn’t solving the problem - only masking it.

Make sure everything else is working as expected. By itself, it’s ok to largely ignore some timeout error unless it’s related to another issue or happening constantly. Make sure your parent relays are within the capacity guidelines and aren’t trying to handle too many connections at once. look out in the log files for “parent relay is busy, backing off”.

It’s important to check the relay buffer directories and making sure they are emptying at a fast enough rate. (relay diagnostics are useful - http://:52311/rd) make sure you check every hop right up to the server itself finishing off by looking at the FillDB log. At the relay level, reasons may be poor bandwidth, however at the server level you may have DB performance issues that are slowing down the FillDB process and causing reports to pile up.

You can also look at increasing the client minimum report interval and the refresh rate. That will reduce the load up the chain. There are other client settings you can change to reduce load so explore those too you think your environment is being pushed.