Bug 1386617

Summary: [ceph-iscsi-ansible]: gateways cannot be configured if cluster name is other than ceph
Product: [Red Hat Storage] Red Hat Storage Console Reporter: Tejas <tchandra>
Component: documentationAssignee: ceph-docs <ceph-docs>
Status: CLOSED EOL QA Contact: ceph-qe-bugs <ceph-qe-bugs>
Severity: urgent Docs Contact: Bara Ancincova <bancinco>
Priority: unspecified    
Version: 2CC: adeza, agunn, aschoen, ceph-eng-bugs, gmeno, hnallurv, kdreyer, mchristi, nthomas, sankarshan
Target Milestone: ---   
Target Release: 2   
Hardware: Unspecified   
OS: Linux   
Whiteboard:
Fixed In Version: ceph-iscsi-ansible-1.3-1.el7scon ceph-iscsi-config-1.3-1.el7cp Doc Type: Known Issue
Doc Text:
.iSCSI gateway setup fails if the cluster name is different than "ceph" The `device-mapper-multipath rbd` path checker currently only supports the default cluster name, which is "ceph". As a consequence, an attempt to set up an iSCSI gateway fails during Logical Unit Number (LUN) creation if the cluster name is different than "ceph". In addition, the `ansible-playbook ceph-iscsi-gw.yml` command returns the following error: ---- Could not find dm multipath device for <image_name>. ---- To work around this problem: . In the `ceph-iscsi-gw` Ansible configuration file: .. Set the `cluster_name` variable to `ceph`. .. Set the `gateway_keyring` variable to `ceph.client.admin.keyring`. . On the `seed_monitor` host, create the following symbolic links: * from `/etc/ceph/<your_cluster_name>.conf` to `/etc/ceph/ceph.conf` + ---- ln -s /etc/ceph/<your_cluster_name>.conf /etc/ceph/ceph.conf ---- * from `/etc/ceph/<your_cluster_name>.client.admin.keyring` to `/etc/ceph/ceph.client.admin.keyring` + ---- ln -s /etc/ceph/<your_cluster_name>.client.admin.keyring \ /etc/ceph/ceph.client.admin.keyring ----
Story Points: ---
Clone Of: Environment:
Last Closed: Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description Tejas 2016-10-19 10:03:18 UTC
Description of problem:
    I have created a ceph cluster named "iscsi1". The task to sync the admin keyring fails, since its looking for the "ceph.client.admin.keyring".
the keyring should be "<cluster>.client.admin.keyring"

ll /etc/ceph/
total 12
-rw-------. 1 ceph ceph  63 Oct 19 01:57 iscsi1.client.admin.keyring
-rw-r--r--. 1 ceph ceph 803 Oct 19 01:57 iscsi1.conf
-rwxr-xr-x. 1 root root  92 Oct 12 12:40 rbdmap



TASK: [ceph-iscsi-gw | sync the admin keyring from the seed monitor to each gateway node] *** 
failed: [havoc -> harvard] => {"cmd": "rsync --delay-updates -F --compress --archive --rsh 'ssh  -S none -o StrictHostKeyChecking=no' --out-format='<<CHANGED>>%i %n%L' \"/etc/ceph/ceph.client.admin.keyring\" \"root@havoc:/etc/ceph/ceph.client.admin.keyring\"", "failed": true, "rc": 23}
msg: rsync: link_stat "/etc/ceph/ceph.client.admin.keyring" failed: No such file or directory (2)
rsync error: some files/attrs were not transferred (see previous errors) (code 23) at main.c(1052) [sender=3.0.9]

failed: [mustang -> harvard] => {"cmd": "rsync --delay-updates -F --compress --archive --rsh 'ssh  -S none -o StrictHostKeyChecking=no' --out-format='<<CHANGED>>%i %n%L' \"/etc/ceph/ceph.client.admin.keyring\" \"root@mustang:/etc/ceph/ceph.client.admin.keyring\"", "failed": true, "rc": 23}
msg: rsync: link_stat "/etc/ceph/ceph.client.admin.keyring" failed: No such file or directory (2)
rsync error: some files/attrs were not transferred (see previous errors) (code 23) at main.c(1052) [sender=3.0.9]

failed: [magna086.ceph.redhat.com -> harvard] => {"cmd": "rsync --delay-updates -F --compress --archive --rsh 'ssh  -S none -o StrictHostKeyChecking=no' --out-format='<<CHANGED>>%i %n%L' \"/etc/ceph/ceph.client.admin.keyring\" \"root.redhat.com:/etc/ceph/ceph.client.admin.keyring\"", "failed": true, "rc": 23}
msg: rsync: link_stat "/etc/ceph/ceph.client.admin.keyring" failed: No such file or directory (2)
rsync error: some files/attrs were not transferred (see previous errors) (code 23) at main.c(1052) [sender=3.0.9]

failed: [magna080.ceph.redhat.com -> harvard] => {"cmd": "rsync --delay-updates -F --compress --archive --rsh 'ssh  -S none -o StrictHostKeyChecking=no' --out-format='<<CHANGED>>%i %n%L' \"/etc/ceph/ceph.client.admin.keyring\" \"root.redhat.com:/etc/ceph/ceph.client.admin.keyring\"", "failed": true, "rc": 23}
msg: rsync: link_stat "/etc/ceph/ceph.client.admin.keyring" failed: No such file or directory (2)
rsync error: some files/attrs were not transferred (see previous errors) (code 23) at main.c(1052) [sender=3.0.9]


FATAL: all hosts have already failed -- aborting

We have a provision to support different names for clusters from Ceph 2.0 onwards.




Version-Release number of selected component (if applicable):
ceph-ansible-1.0.5-36.el7scon.noarch
ceph-iscsi-ansible-1.2-1.el7scon.noarch

How reproducible:
Always

Steps to Reproduce:
1. on a cluster which has a user defined name, try to create the iscsi gateways

Comment 4 Paul Cuzner 2016-10-19 21:42:27 UTC
Changing the cluster name not only impacts the keys, but also the default interaction to the cluster - since ceph.conf becomes <cluster_name>.conf

Good catch - the fix is likely to touch a number of modules.

Comment 5 Paul Cuzner 2016-10-21 05:38:51 UTC
Resolved in v1.3 - please retest with this version

Comment 6 Paul Cuzner 2016-10-25 00:26:01 UTC
Please review the additional variables set in group_vars file that define the cluster name and admin keyring to use.

Comment 12 Mike Christie 2016-11-04 17:39:24 UTC
We are only going to doc this for 2.1. Bug info and workaround added to Doc Text.

Comment 18 Shubhendu Tripathi 2018-11-19 05:44:11 UTC
This product is EOL now