This article describes an issue where Host Checker reevaluation fails in Mac OS X machines with dsAccessService crash
Problem or Goal
Host Checker reevaluation fails in Mac OS X machines with dsAccessService crash
When the issue triggers the Pulse Client connection will be torn down and reauthentication is needed to connect to back PCS gateway
This can also be validated by saving the Pulse Client debug logs with detailed level
Open Pulse client Navigate to File >> Logs >> Log level >> make it to detailed . Again Navigate to File >> Logs >>Click on " Save as " and save the logs
Unzip the file which you saved in the above step and Navigate to DiagnosticReports folder and see if you can see a crash with a namefor example " dsAccessService_2021-01-20-081137_Test-Mac.crash "
You will also see the below snippet from user access logs when this issue is seen
info - [x.x.x.x] - Default Network::UserABC(Domain)[Domain](25f4daacc5) - 2021/01/21 15:10:56 - System process detected a Host Checker time out on host x.x.x.x for user 'UserABC' (last update at 2021-01-21 14.50.41 +0000 GMT).
info - [x.x.x.x] - Default Network::UserABC(Domain)[Domain] - 2021/01/21 15:10:56 - Closed connection to x.x.x.x after 1215 seconds, with 1747901 bytes read and 328553 bytes written
This is caused due to the OPSWAT library is crashing while reevaluation is triggered in the endpoint machine
To resolve the issue upgrade the ESAP package to 3.7.4 and above.
As a workaround, we can increase the Host checker reevaluation time to a bigger number
Navigate to Authentication >> Endpoint Security >> Host checker and change the value to a bigger number under Perform check every --- minutes field
In the above screenshot, it has been set to 90 Minutes so the Host checker reevaluation will be triggered after 90 minutes of the successful Pulse Client connection
This is a Global configuration and not specific to any single Realm
So doing this change will affect all the realms configured in the PCS gateway