You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Version: 3.13.7-management-alpine (amd64)
sha256:c691b282f98d62beea4b1fb293c8b12c76208a489c3656c234c8a865043e8249 (latest for the tag as of 2024-12-19) System: Linux 6.1.0-28-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.119-1 (2024-11-22) x86_64 GNU/Linux Observed behavior: The number of file descriptors is limited to 1024 Expected behavior: The number of file descriptors corresponds to the limits defined by the docker daemon and/or the system
Details:
The number of file descriptors is limited to 1024 even though the process runs with the same limits as the system (2^20 file descriptors).
At server start, the following warning appears.
2024-12-19 16:06:07.000718+01:00 [warning] <0.254.0> Recovering 1603 queues, available file handles: 927. Please increase max open file handles limit to at least 1603!
The boot process doesn't finish (I interrupted it after more than an hour) and the service doesn't become available.
I have tried setting ulimit manually beforehand, and also tried defining the ulimits directly in the docker daemon config files to no avail.
Resolution attemps:
I have another cluster using the same 3.13.7-management-alpine (amd64) version. However, it was deployed earlier, the digest differs (sha256:a4c3e1b9ec43fddb486e3560d4dea0ac1d520d789b3fa22646850d05bd9c0e1f) and the problem does not occur
The problem does not occur when going back to version 3.13.6-management-alpine (sha256:611107e29cce05c2acd968325d5dcbde7e2fee404970f1ead75fdb22be2821b3)
The text was updated successfully, but these errors were encountered:
Architect-unix-Boursorama
changed the title
BUG: Not enough file descriptors to handle files
BUG: Not enough file descriptors to handle queues
Dec 19, 2024
Version: 3.13.7-management-alpine (amd64)
sha256:c691b282f98d62beea4b1fb293c8b12c76208a489c3656c234c8a865043e8249 (latest for the tag as of 2024-12-19)
System: Linux 6.1.0-28-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.119-1 (2024-11-22) x86_64 GNU/Linux
Observed behavior: The number of file descriptors is limited to 1024
Expected behavior: The number of file descriptors corresponds to the limits defined by the docker daemon and/or the system
Details:
The number of file descriptors is limited to 1024 even though the process runs with the same limits as the system (2^20 file descriptors).
At server start, the following warning appears.
2024-12-19 16:06:07.000718+01:00 [warning] <0.254.0> Recovering 1603 queues, available file handles: 927. Please increase max open file handles limit to at least 1603!
The boot process doesn't finish (I interrupted it after more than an hour) and the service doesn't become available.
I have tried setting ulimit manually beforehand, and also tried defining the ulimits directly in the docker daemon config files to no avail.
Resolution attemps:
I have another cluster using the same 3.13.7-management-alpine (amd64) version. However, it was deployed earlier, the digest differs (sha256:a4c3e1b9ec43fddb486e3560d4dea0ac1d520d789b3fa22646850d05bd9c0e1f) and the problem does not occur
The problem does not occur when going back to version 3.13.6-management-alpine (sha256:611107e29cce05c2acd968325d5dcbde7e2fee404970f1ead75fdb22be2821b3)
The text was updated successfully, but these errors were encountered: