This issue occurs when the IPMI System Event log file is full, which triggers the alarm that monitors the log file.
Please refer "Host IPMI system event log status" alarm in vCenter Server (1033725) to clear the IPMI System Event log file.
"A general system error occurred: Retrieve IPMI SEL request to host failed" is quite common observed error in vSphere 6.5 together with:
Take a look at the log settings for vpxa (/etc/vmware/vpxa/vpxa.cfg). I have found on my 6.5 servers, that the vpxa log-level was set to verbose. Thus the vpxa log fills quite quickly and if it is full you get those errors. Restarting vpxa is one method, but that is only temporary as you are hit by the next full log in foreseeable future.
I set the log-level of vpxa to info and restarted vpxa. Since this change all seems to work.
Maybe this helps in your situation, too,
Here's what I found in the VPXA.CFG
For me it was sufficient to set the general log level in the second row of your file to <level>info</level> instead of verbose. You can further set the loglevel for memory (<memoryLevel>...</memoryLevel>) to info, as that is set to verbose now.
Probably the verbose setting is now default for new installations, which can lead to the problems described by you.
"A general system error occurred: Retrieve IPMI SEL request to host failed" - looks like this has been resolved in the latest update of ESXi 6.5 (build 7967591).
I've done upgrade using DellEMC-ESXi-6.5U1-7967591-A10 ISO (VMware ESXi 6.5 U1 Driver Details | Dell Australia), and now it works as expected:
Previously, with DellEMC-ESXi-6.5U1-7388607-A07 it was still generating an error message:
Hope this information helps.