Bypass RBLs After Delay Threshold?
Question asked by Jay Dubb - 2/8/2022 at 9:36 AM
We've seen some spool backups lately, and have traced it to sluggish RBLs, specifically SORBS.  Latency to other RBLs is normal (sub-100 ms) but there have been several instances of SORBS exceeding 6,000 ms.  When spool volumes are running around 1500 messages, a delay of 6 seconds per message really backs things up.

Is there a setting that I'm missing, that will cause an RBL to be bypassed if latency exceeds a certain (admin configurable) threshold?  For example, if an RBL takes longer than 1,000 ms to respond, bypass for this message and move on.

1 Reply

Reply to Thread
Scarab Replied
There isn't a setting, at least that I'm aware of, but there definitely needs to be. We also have had occasions where an RBL is entirely unresponsive, or when we have exceeded hourly thresholds allowed by a RBL where rather than giving an error response that  RBL just doesn't respond, and in both cases SM simply doesn't timeout causing thousands upon thousands of messages to just pile up in the Spool waiting to be processed.

When this happens we just have to find the RBL causing it, disable it, and restart the Spool. There really should be a configurable timeout setting for antispam checks.

Reply to Thread