Bug #6272
closeddpdk: big mempool leads to an error with suricatasc unix socket
Updated by Victor Julien 8 months ago
- Target version changed from 7.0.1 to 7.0.2
Updated by Lukas Sismis 8 months ago
Analyzed this the other day and I can reproduce this on Mellanox but not on Intel cards (i40e/ixgbe).
So far the workaround was to allocate 1 GB hugepages instead of 2 MB ones.
Updated by Lukas Sismis 8 months ago
Now in a PR mix - https://github.com/OISF/suricata/pull/9519
Updated by Lukas Sismis 7 months ago
- Status changed from New to In Review
Updated by Victor Julien 7 months ago
- Target version changed from 7.0.2 to 7.0.3
Updated by Lukas Sismis 7 months ago
Error message:
Perf: threads: Setting prio 0 for thread "US", thread id 600975 [TmThreadSetupOptions:tm-threads.c:884]
Notice: threads: Threads created -> W: 36 FM: 1 FR: 1 Engine started. [TmThreadWaitOnThreadRunning:tm-threads.c:1893]
Info: dpdk: 10735 of 13312 of hugepages are free - number of hugepages can be lowered to e.g. 2964 [MemInfoEvaluateHugepages:util-dpdk.c:139]
^CNotice: suricata: Signal Received. Stopping engine. [SuricataMainLoop:suricata.c:2816]
Error: unix-manager: Command server: select() fatal error: Bad file descriptor [UnixMain:unix-manager.c:648]
Error: unix-manager: Fatal error on unix socket [UnixManager:unix-manager.c:1163]
...
Steps on how to reproduce:
https://forum.suricata.io/t/suricata-version-7-0-1-dev-dpdk-rx-tx-descriptors-and-suricatasc-fail/3823/8?u=lukashino