Troubleshoot an issue where a new RGW deployment is failing because of an address already in use. This bug corrupts the health of the Ceph cluster.
Return the Ceph cluster to a healthy state by resolving issues with failing deployment.
This procedure requires admin privileges.
See Collect Information about the Ceph Cluster for more information on how to interpret the output of the Ceph commands used in this procedure.
Log on to a master node or ncn-s001/2/3 to run the following commands.
Observe that a Ceph RGW deployment is failing to deploy because the address is already in use.
Use the cephadm logs to see if this error is occurring. The following command is following the cephadm logs. It may take a minute or two for the Ceph to attempt to deploy RGW and for the error to appear.
ceph -W cephadm
Example output:
cluster:
id: 56cdd77c-7184-11ef-9f67-42010afc0104
health: HEALTH_WARN
Failed to place 2 daemon(s)
services:
mon: 3 daemons, quorum ncn-s001,ncn-s002,ncn-s003 (age 7d)
mgr: ncn-s002.mneqjy(active, since 7d), standbys: ncn-s001.iwikun, ncn-s003.pkabeg
mds: 2/2 daemons up, 3 standby, 1 hot standby
osd: 18 osds: 18 up (since 7d), 18 in (since 7d)
rgw: 3 daemons active (3 hosts, 1 zones)
data:
volumes: 2/2 healthy
pools: 15 pools, 625 pgs
objects: 69.74k objects, 123 GiB
usage: 260 GiB used, 8.7 TiB / 9.0 TiB avail
pgs: 625 active+clean
io:
client: 8.4 KiB/s rd, 1.8 MiB/s wr, 5 op/s rd, 176 op/s wr
progress:
Updating rgw.site1.zone1 deployment (+2 -> 2) (2s)
[============================]
2024-09-20T19:04:56.337199+0000 mgr.ncn-s002.mneqjy [INF] Removing key for client.rgw.site1.zone1.ncn-s002.mgujxl
2024-09-20T19:04:56.371965+0000 mgr.ncn-s002.mneqjy [ERR] Failed while placing rgw.site1.zone1.ncn-s002.mgujxl on ncn-s002: cephadm exited with an error code: 1, stderr: Non-zero exit code 125 from /usr/bin/podman container inspect --format {{.State.Status}} ceph-56cdd77c-7184-11ef-9f67-42010afc0104-rgw-site1-zone1-ncn-s002-mgujxl
/usr/bin/podman: stderr Error: no such container ceph-56cdd77c-7184-11ef-9f67-42010afc0104-rgw-site1-zone1-ncn-s002-mgujxl
Non-zero exit code 125 from /usr/bin/podman container inspect --format {{.State.Status}} ceph-56cdd77c-7184-11ef-9f67-42010afc0104-rgw.site1.zone1.ncn-s002.mgujxl
/usr/bin/podman: stderr Error: no such container ceph-56cdd77c-7184-11ef-9f67-42010afc0104-rgw.site1.zone1.ncn-s002.mgujxl
Deploy daemon rgw.site1.zone1.ncn-s002.mgujxl ...
Verifying port 8080 ...
Cannot bind to IP 0.0.0.0 port 8080: [Errno 98] Address already in use
ERROR: TCP Port(s) '8080' required for rgw already in use
2024-09-20T19:04:56.410687+0000 mgr.ncn-s002.mneqjy [INF] Deploying daemon rgw.site1.zone1.ncn-s001.rdtynd on ncn-s001
2024-09-20T19:04:57.536533+0000 mgr.ncn-s002.mneqjy [INF] Removing key for client.rgw.site1.zone1.ncn-s001.rdtynd
2024-09-20T19:04:57.576404+0000 mgr.ncn-s002.mneqjy [ERR] Failed while placing rgw.site1.zone1.ncn-s001.rdtynd on ncn-s001: cephadm exited with an error code: 1, stderr: Non-zero exit code 125 from /usr/bin/podman container inspect --format {{.State.Status}} ceph-56cdd77c-7184-11ef-9f67-42010afc0104-rgw-site1-zone1-ncn-s001-rdtynd
/usr/bin/podman: stderr Error: no such container ceph-56cdd77c-7184-11ef-9f67-42010afc0104-rgw-site1-zone1-ncn-s001-rdtynd
Non-zero exit code 125 from /usr/bin/podman container inspect --format {{.State.Status}} ceph-56cdd77c-7184-11ef-9f67-42010afc0104-rgw.site1.zone1.ncn-s001.rdtynd
/usr/bin/podman: stderr Error: no such container ceph-56cdd77c-7184-11ef-9f67-42010afc0104-rgw.site1.zone1.ncn-s001.rdtynd
Deploy daemon rgw.site1.zone1.ncn-s001.rdtynd ...
Verifying port 8080 ...
Cannot bind to IP 0.0.0.0 port 8080: [Errno 98] Address already in use
ERROR: TCP Port(s) '8080' required for rgw already in use
Note the name of the daemon attempting to deploy for the next step, in this case rgw.site1.zone1.
Verify that the running daemon has a different name than the one attempting to deploy
Check the names of the running daemons with the following command.
ceph orch ps --daemon_type rgw
Example output:
NAME HOST PORTS STATUS REFRESHED AGE MEM USE MEM LIM VERSION IMAGE ID CONTAINER ID
rgw.site1.ncn-s001.xoaosp ncn-s001 *:8080 running (6d) 7m ago 6d 516M - 17.2.6 6eebe3129025 fdd8842a0b16
rgw.site1.ncn-s002.qsibkp ncn-s002 *:8080 running (6d) 7m ago 6d 478M - 17.2.6 6eebe3129025 86bd2327ab81
rgw.site1.ncn-s003.hwiydc ncn-s003 *:8080 running (4d) 3m ago 4d 412M - 17.2.6 6eebe3129025 fb89e37f1ec8
In this example, observe that the name of the running RGW deployment is rgw.site1. This is different than the RGW deployment that is trying to be deployed above which has the name rgw.site1.zone1.
If the error is occuring and the running daemon has a different name than the one attempting to deploy continue on to resolve the two RGW deployments.
Remove the conflicting daemon.
Note: The RGW deployment in CSM should be named rgw.site1. The deployment that should be removed should have a name other than rgw.site1. In this example, the deployment that should be removed is rgw.site1.zone1.
This command will remove the conflicting daemon and should restore the Ceph cluster to a healthy state.
ceph orch rm rgw.site1.zone1
Example output:
Removed service rgw.site1.zone1
Check the health of the Ceph cluster on one of the manager nodes.
Rerun the command to check the Ceph admin tool.
ceph -W cephadm
Healthy example output:
cluster:
id: 56cdd77c-7184-11ef-9f67-42010afc0104
health: HEALTH_OK
services:
mon: 3 daemons, quorum ncn-s001,ncn-s002,ncn-s003 (age 11d)
mgr: ncn-s002.mneqjy(active, since 11d), standbys: ncn-s001.iwikun, ncn-s003.pkabeg
mds: 2/2 daemons up, 3 standby, 1 hot standby
osd: 18 osds: 18 up (since 11d), 18 in (since 11d)
rgw: 3 daemons active (3 hosts, 1 zones)
data:
volumes: 2/2 healthy
pools: 15 pools, 625 pgs
objects: 74.42k objects, 137 GiB
usage: 277 GiB used, 8.7 TiB / 9.0 TiB avail
pgs: 625 active+clean
io:
client: 12 KiB/s rd, 3.9 MiB/s wr, 3 op/s rd, 374 op/s wr