Queue Manager restarts when AgentGroups configured to collect statistics
| Issue date | June 5, 2020 | |
| Observed behavior | Queue Manager unexpectedly restarts when several AgentGroups are configured in TIAL Avaya TSAPI environments. Other environments could also experience a restart as well. | |
| Products and versions affected | Mindful Enterprise Callback 9.0 | |
| Components affected | cti_connector_host | |
| Applications affected | Queue Manager | |
| Severity | Major | |
| Root cause | Multiple TIAL commands simultaneously being sent from Queue Manager to the cti_connector_host cause the host to read in memory for the wrong request; it recognizes this immediately and shuts down and shuts down Queue Manager. This can happen when several (seen with 20+) Agent Groups are configured for Queue Manager to collect real time statistics in a TIAL Avaya TSAPI environment. After a review of the codebase, it has been determined that there is a possibility of this happening for all TIAL integrations under different circumstances, however it has not been observed in our load testing prior to release. | |
| Resolution | The hotfix allows commands to be processed one at a time instead of being sent before the other one is completely processed.Additional details on the available hotfix can be found in the VHT Callback 9.0.1 hotfixdocumentation. | |
| Expected availability | Enterprise Callback 9.0.1, released June 12, 2020. | |
| Workarounds | If AgentGroups are configured for real time statistics, use the RTAgentGroups instead and configure the Real-Time Adapter to collect statistics. | |
| Defects related to this issue | 173069417 | Make TIAL commands thread safe |
