TORQ框架中tp故障恢复后rdb未自动重新订阅的问题咨询

Hey there, let's break down this TORQ issue you're running into.

First off, TORQ does have built-in functionality for automatic re-subscription after a component recovery, but there are a few common factors that might be preventing it from working in your case:

• Check your RDB configuration: Make sure your rdb instance has the appropriate auto-reconnect/subscription retry parameters enabled. Depending on your TORQ deployment setup, you might need to add flags like -reconnect or configure subscription retry intervals in the rdb startup script or config file. If this setting is missing, rdb won't attempt to re-subscribe on its own after a tp outage.
• Review recovery event signaling: When your tp was automatically brought back up, check if the recovery process properly sent a subscription trigger event to the rdb. Sometimes automated recovery workflows don't hook into TORQ's internal event bus correctly, so rdb never gets notified that tp is back online and ready for subscriptions. Dig into both tp and rdb logs around the recovery time to spot any missing event signals.
• Test manual subscription triggers: Try using TORQ's built-in rdb commands (like \sub to re-initiate subscriptions) to see if you can manually get rdb connected to the recovered tp. If this works, it confirms rdb's subscription functionality is intact, and the issue is with the automatic trigger logic.

As for whether you need to bounce rdb and wdb in these cases: A manual restart will certainly resolve the immediate issue, but it's not a sustainable fix. Your goal should be to get the automatic re-subscription working by fixing the configuration or event signaling gap. That said, if you're in a pinch and need to restore functionality quickly, restarting those components is a valid temporary workaround.

备注：内容来源于stack exchange，提问作者flavio