This error should have been caught as an escalation failure:
"Escalation": [
{"failed": [
{"failure_reason": "liveagent bot offline or disabled"} <- DF.app
]}
NOTE: I'm seeing this error correlated with sessions that are being closed by the df_bot due to no appName/token. However, it appears that the visitor is able to type into the composer, which triggers the DF fallback and attempts an escalation, which fails with this error. Therefore, this particular error seems to be the signature of a composer bug or race condition.
Our event logging will show an attempted escalation in this case, but not a failed escalation.
We should fix this event and consider adding another property to indicate something like 'composer_disabled_failure'
example:
Rocke.Chat Livechat ERROR: Failed to handover {"error":"","errorMessage":"Unable to reach Liveagent bot, it may be offline or disabled.","dialogflow_SessionID":"e5AnteBt2ZPeAFsRA","visitorDetails":{"id":"j38cTqdfPW6bvsKiF","token":"3c05e4fc5d17097ce5673712a2fdbe5212bff7b95d57b5b58cfcd65e3a5aefbf"},"targetDepartment":"hdgncTTn5YBrxyLLc"}
This error should have been caught as an escalation failure:
NOTE: I'm seeing this error correlated with sessions that are being closed by the df_bot due to no appName/token. However, it appears that the visitor is able to type into the composer, which triggers the DF fallback and attempts an escalation, which fails with this error. Therefore, this particular error seems to be the signature of a composer bug or race condition.
Our event logging will show an attempted escalation in this case, but not a failed escalation.
We should fix this event and consider adding another property to indicate something like 'composer_disabled_failure'
example: