liaoxin01 commented on code in PR #64301:
URL: https://github.com/apache/doris/pull/64301#discussion_r3443192741
##########
fe/fe-common/src/main/java/org/apache/doris/common/Config.java:
##########
@@ -1193,6 +1193,12 @@ public class Config extends ConfigBase {
@ConfField(mutable = true, masterOnly = true)
public static int streaming_task_timeout_multiplier = 10;
+ /**
+ * streaming task min timeout second.
+ */
+ @ConfField(mutable = true, masterOnly = true)
+ public static int streaming_task_min_timeout_sec = 300;
Review Comment:
[P2] Is a 300-second minimum justified here? With the default
`max_interval=10s` and multiplier `10`, this raises the effective no-progress
timeout from 100 seconds to 300 seconds, and also lets the CDC
startup/WAL-search phase wait up to 150 seconds because it uses half of the
task timeout. Since this PR already renews the deadline when `scannedRows`
advances, legitimate long-running snapshot work should not require such a large
fixed floor. Please provide production evidence for the five-minute minimum, or
reduce it to a smaller value such as 60-120 seconds / derive it from the
heartbeat and RPC timing bounds so genuine stalls recover promptly.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
