Let me share few findings of our support and dev teams on this with you.
Endpoint 9 started to use EPNS instead of DNS requests to check for license changes.
Endpoints 8 and below didn’t report an issue if this check failed.
When it comes to issues reported on ESET PROTECT Virtual appliance
The issue seems to be the default setting of the following module "reqtimeout_module" which is used/loaded only on VA (this module provides a way to set timeouts and minimum data rates for receiving requests).
The issue should not be present on the Windows version of the Apache HTTP PROXY because this module is not loaded there. The workaround could be (1) disabling this module on the VA or (2) setting the appropriate values. We are not sure what is the preferred way and how it might affect other services, as it was not fully tested yet.
To disable limits (i.e. "(2) setting the appropriate values"):
set "RequestReadTimeout header=0 body=0"
in the newly created(in /etc/httpd/conf.d/) configuration file "reqtimeout.conf" with settings mentioned above
and of course the file has to be included in used configuration "IncludeOptional conf.d/reqtimeout.conf" in the "/etc/httpd/conf/httpd.conf"
We are expecting some official solution in the following days.
If the customers do not use the Apache HTTP Proxy on VA, enable the Direct Cloud advanced logging, reproduce the issue, collect the logs by ELC and open a ticket for our support teams to check.
The configuration option for App status and Notification of "Eset Push Notification Service server cannot be reached" state will be added (P_EESW-8067)
The "Do not remind me again" from EPNS app status doesn't work issue is tracked to be fixed (P_EESW-8048)
Regards, Peter
