chovy-3012 opened a new issue, #13306:
URL: https://github.com/apache/seatunnel/issues/13306

   
   ### Search before asking
   
   - [x] I had searched in the 
[issues](https://github.com/apache/seatunnel/issues?q=is%3Aissue+label%3A%22bug%22)
 and found no similar issues.
   
   ### What happened
   
   When the Doris FE that a source job talks to is slow to respond or 
unreachable, the job is not bounded by the configured request timeouts. The 
docs describe `doris.request.connect.timeout.ms` as the connection timeout for 
requests sent to FE or BE, but the FE requests the Doris source sends are 
effectively unbounded.
   
   Measured with both timeouts set to 3000 ms and `doris.request.retries = 1`:
   
   - Unreachable FE address: the job logged the FE request and then produced no 
Doris-related output for 60 seconds — no retry, no error, no failure — and had 
to be killed externally.
   - Slow FE (answered only after 15 s): the job waited the full 15 s and 
accepted the late response, so `doris.request.read.timeout.ms` had no effect 
either.
   
   ### SeaTunnel Version
   
   3.0.0 (the same behavior is present in 2.3.10–2.3.13 as well).
   
   ### SeaTunnel Config
   
   Precondition: a Doris FE reachable on its query port (default 9030) with 
database `demo` and table `tbl` — job startup and the metadata fetch succeed — 
while the address configured in `fenodes` is unreachable (connection attempts 
there are neither accepted nor refused).
   
   ```hocon
   env {
     execution.parallelism = 1
   }
   
   source {
     Doris {
       fenodes = "127.0.0.1:18081"
       username = "root"
       password = ""
       database = "demo"
       table = "tbl"
   
       doris.request.connect.timeout.ms = 3000
       doris.request.read.timeout.ms = 3000
       doris.request.retries = 1
     }
   }
   
   sink {
     Console {}
   }
   ```
   
   For the read-timeout case the config is the same, with `fenodes` pointing at 
a FE that answers only after ~15 s (the second log excerpt below is from that 
run, against its own `SlowDB`.`SlowTable`).
   
   ### Running Command
   
   ```shell
   sh bin/seatunnel.sh --config /path/to/repro.conf -e local
   ```
   
   ### Error Exception
   
   No exception is thrown — the job just keeps waiting.
   
   Unreachable FE with `doris.request.connect.timeout.ms = 3000`, real log 
output:
   
   ```log
   2026-11-17 18:30:51,199 INFO  [s.s.DorisSourceSplitEnumerator] 
[BlockingWorker-TaskGroupLocation{jobId=1175020216766169089, pipelineId=1, 
taskGroupId=1}] - Send request to Doris FE 
'http://127.0.0.1:18081/api/demo/tbl/_query_plan' with user 'root'.
   2026-11-17 18:31:50,997 WARN  [o.a.s.e.c.j.JobMetricsRunner  ] 
[job-metrics-runner-1175020216766169089] - Failed to get job metrics summary, 
it maybe first-run
   ```
   
   These two lines are 60 seconds apart. In between, the job logged nothing 
about the FE request — no retry, no error, no failure. It was still blocked in 
the connection attempt and had to be killed externally.
   
   Slow FE with `doris.request.read.timeout.ms = 3000` (FE answered after 15 
s), real log output:
   
   ```log
   2026-11-17 18:30:32,511 INFO  [s.s.DorisSourceSplitEnumerator] 
[BlockingWorker-TaskGroupLocation{jobId=1175020138223632385, pipelineId=1, 
taskGroupId=1}] - Send request to Doris FE 
'http://127.0.0.1:18080/api/SlowDB/SlowTable/_query_plan' with user 'root'.
   2026-11-17 18:30:47,920 INFO  [s.c.s.s.c.ClientExecuteCommand] [main] - 
Closed HazelcastInstance ......
   ```
   
   The request was sent at 18:30:32 and the job only shut down at 18:30:47 — it 
waited the full 15 s for the answer, accepted the late response and finished 
successfully. The 3000 ms read timeout never triggered.
   
   | Scenario | Config | Expected | Actual |
   |---|---|---|---|
   | FE address unreachable (connection attempts neither accepted nor refused) 
| `doris.request.connect.timeout.ms = 3000` | The request fails within ~3 s and 
the job fails fast instead of hanging | No Doris-related output for 60 s — no 
retry, no error, no failure; the job had to be killed externally |
   | FE answers only after ~15 s | `doris.request.read.timeout.ms = 3000` | The 
request times out after ~3 s | The job waited the full 15 s, accepted the late 
response and finished successfully |
   
   ### Zeta or Flink or Spark Version
   
   Zeta (local mode)
   
   ### Java or Scala Version
   
   Java 11
   
   ### Are you willing to submit PR?
   
   - [x] Yes I am willing to submit a PR!
   
   ### Code of Conduct
   
   - [x] I agree to follow this project's [Code of 
Conduct](https://www.apache.org/foundation/policies/conduct)
   


-- 
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]

Reply via email to