Hello,
in a NFS cluster with ingress IP/service after migrating the NFS backend
service/daemon to another Ceph node (host pinning)
the (new) deployed NFS daemon/container uses the ingress
port 2049 (default) as backend port instead the port (12049) used at NFS cluster
creation.
When NFS backend gets deployed to a node where NFS haproxy daemon is running
then backend fails to start due to "Address already in use".
Is there a way to workaround this situation?
Message is:
[WARN] CEPHADM_DAEMON_PLACE_FAIL: Failed to place 1 daemon(s)
Failed while placing nfs.fmi-ceph-nfs.0.2.ceph02.busmox on ceph02:
cephadm exited with an error code: 1,
stderr:Non-zero exit code 1 from /usr/bin/docker container inspect --format
{{.State.Status}}
ceph-f54eea86-265a-11eb-a5d0-457857ba5742-nfs-fmi-ceph-nfs-0-2-ceph02-busmox
/usr/bin/docker: stdout /usr/bin/docker: stderr Error: No such container:
ceph-f54eea86-265a-11eb-a5d0-457857ba5742-nfs-fmi-ceph-nfs-0-2-ceph02-busmox
Non-zero exit code 1 from /usr/bin/docker container inspect --format
{{.State.Status}}
ceph-f54eea86-265a-11eb-a5d0-457857ba5742-nfs.fmi-ceph-nfs.0.2.ceph02.busmox
/usr/bin/docker: stdout /usr/bin/docker: stderr Error: No such container:
ceph-f54eea86-265a-11eb-a5d0-457857ba5742-nfs.fmi-ceph-nfs.0.2.ceph02.busmox
Deploy daemon nfs.fmi-ceph-nfs.0.2.ceph02.busmox ...
Verifying port 2049 ...
Cannot bind to IP 0.0.0.0 port 2049: [Errno 98] Address already in use
ERROR: TCP Port(s) '2049' required for nfs already in use
Thanks & best regards
Uwe
Details:
# cephadm shell ceph version
ceph version 16.2.9 (4c3647a322c0ff5a1dd2344e039859dcbd28c830) pacific (stable)
# cephadm shell ceph nfs cluster create fmi-ceph-nfs --ingress --virtual-ip
10.149.13.49/22
# cephadm shell ceph nfs cluster info
{
"fmi-ceph-nfs": {
"virtual_ip": "10.149.13.49",
"backend": [
{
"hostname": "ceph01",
"ip": "10.149.12.21",
===> "port": 12049
}
],
"port": 2049,
"monitor_port": 9049
}
}
# ceph orch ps --service-name ingress.nfs.fmi-ceph-nfs
NAME HOST PORTS STATUS REFRESHED AGE MEM USE MEM LIM VERSION IMAGE ID CONTAINER ID
haproxy.nfs.fmi-ceph-nfs.ceph01.qvltza ceph01 *:2049,9049 running (52s) 26s ago 52s 14.2M - 2.3.17-d1c9119 e85424b0d443 1e7823bda90c
haproxy.nfs.fmi-ceph-nfs.ceph02.ibzlmt ceph02 *:2049,9049 running (49s) 27s ago 49s 14.0M - 2.3.17-d1c9119 e85424b0d443 2fb82d10011c
keepalived.nfs.fmi-ceph-nfs.ceph01.auiqoj ceph01 running (47s) 26s ago 47s 5984k - 2.0.5 073e0c3cd1b9 0884af833dcc
keepalived.nfs.fmi-ceph-nfs.ceph02.kwvgfh ceph02 running (45s) 27s ago 45s 8527k - 2.0.5 073e0c3cd1b9 949afa050cee
# ceph orch ps --service-name nfs.fmi-ceph-nfs
NAME HOST PORTS STATUS REFRESHED AGE MEM USE MEM LIM VERSION IMAGE ID CONTAINER ID
nfs.fmi-ceph-nfs.0.0.ceph01.qeiwle ceph01 *:12049 running (93s) 65s ago 93s 43.6M - 3.5 3520ead5eb19 a659099b7e11
# now host pinning nfs.fmi-ceph-nfs to node ceph02
# cephadm shell ceph nfs cluster info
Using recent ceph image
ceph/ceph@sha256:056637972a107df4096f10951e4216b21fcd8ae0b9fb4552e628d35df3f61139
{
"fmi-ceph-nfs": {
"virtual_ip": "10.149.13.49",
"backend": [
{
"hostname": "ceph02",
"ip": "10.149.12.22",
===> "port": 2049
}
],
"port": 2049,
"monitor_port": 9049
}
}
# ===> Cannot bind to IP 0.0.0.0 port 2049: [Errno 98] Address already in use
ERROR: TCP Port(s) '2049' required for nfs already in use
--
Uwe Richter
Friedrich-Schiller-Universität Jena,
Fakultät für Mathematik und Informatik, KSZ
Ernst-Abbe-Platz 2, Zi. 3418, D-07740 Jena
mailto:uwe.richter@xxxxxxxxxxx
_______________________________________________
Dev mailing list -- dev@xxxxxxx
To unsubscribe send an email to dev-leave@xxxxxxx
