Created attachment 942931 [details] Gluster and volume logs Description of problem: Creating a replicated volume with two bricks (RDO test day Cinder), setting auth.allow * -> # gluster volume set tshefi-cin-rep auth.allow* volume set: failed: option : mypol.pp does not Did you mean auth.allow? Version-Release number of selected component (if applicable): RHEL7 glusterfs-libs-3.5.2-1.el7.x86_64 glusterfs-server-3.5.2-1.el7.x86_64 glusterfs-api-3.5.2-1.el7.x86_64 glusterfs-fuse-3.5.2-1.el7.x86_64 glusterfs-cli-3.5.2-1.el7.x86_64 glusterfs-3.5.2-1.el7.x86_64 How reproducible: Not sure done only once Steps to Reproduce: 1. Create a replicated volume: #gluster volume create tshefi-cin-rep replica 2 transport tcp 10.35.163.139:/export/tshefi-cin-rep 10.35.163.79:/export/tshefi-cin-rep/ gluster volume set tshefi-cin-rep storage.owner-uid 165 gluster volume set tshefi-cin-rep storage.owner-gid 165 2. Set volume options (my case Cinder) gluster volume set tshefi-cin-rep root-squash enable gluster volume set tshefi-cin-rep auth.allow* volume set: failed: option : mypol.pp does not exist 3. gluster volume info tshefi-cin-rep Volume Name: tshefi-cin-rep Type: Replicate Volume ID: 702f7ffe-b34b-40d7-b40c-0a07cf4d4ac6 Status: Created Number of Bricks: 1 x 2 = 2 Transport-type: tcp Bricks: Brick1: 10.35.163.139:/export/tshefi-cin-rep Brick2: 10.35.163.79:/export/tshefi-cin-rep Options Reconfigured: storage.owner-gid: 165 storage.owner-uid: 165 server.root-squash: enable 4. I have also tired to use command after starting volume, still no luck gluster volume start tshefi-cin-rep volume start: tshefi-cin-rep: success gluster volume set tshefi-cin-rep auth.allow * Usage: volume set <VOLNAME> <KEY> <VALUE> Actual results: Set volume auth.allow * doesn't work. Expected results: Volume set auth.allow * should work and show up with volume info.
So I figured I'd try with specific IP and it works, problem only with "*" gluster volume set tshefi-cin-rep auth.allow 10.35.160.137 volume set: success [root@dhcp163-130 glusterfs]# gluster volume info tshefi-cin-rep Volume Name: tshefi-cin-rep Type: Replicate Volume ID: 702f7ffe-b34b-40d7-b40c-0a07cf4d4ac6 Status: Started Number of Bricks: 1 x 2 = 2 Transport-type: tcp Bricks: Brick1: 10.35.163.139:/export/tshefi-cin-rep Brick2: 10.35.163.79:/export/tshefi-cin-rep Options Reconfigured: auth.allow: 10.35.160.137 storage.owner-gid: 165 storage.owner-uid: 165 server.root-squash: enable
I think this is the how the shell behaves. When you mention *, it just refers to all files in current dir. You can use "*" to set the value (i.e) you should use volume set in the following way, gluster volume set tshefi-cin-rep auth.allow "*" I don't think this issue can be fixed. Kaushal, any thoughts ?
Here is the demonstration of the argument that I have posted in comment3 [root@ ~]# gluster volume set r auth.allow "*" volume set: success [root@host ~]# gluster volume info r Volume Name: r Type: Distribute Volume ID: 7074bf6a-22c3-4a34-87a6-a44cf6372064 Status: Created Number of Bricks: 1 Transport-type: tcp Bricks: Brick1: host1.example.com:/tmp/1 Options Reconfigured: auth.allow: *
As per # comment 3 this is an expected behaviour, you would need to use quotes to pass * to the shell otherwise it will stand for all and command will fail. Closing this bug.