Description of problem: ======================= Had a 1*3 volume, executed 'gluster-block create <volname>/<blockname> ha 3 <IP1>,<IP2>,<IP1> <size>'. Please note that 'IP1' was given twice. The first time, VM core was generated [Bug 1473162]. Other attempts to reproduce it (and tried it thrice, on different nodes) resulted in VM hang. I am assuming the code checks only for the 'number' of addresses given, and has no check for the validity/duplicity of the addresses. I still don't understand the reason of hang though. But it would be good to check the sanity of IPs at the CLI level itself. I am guessing that this might not actually be hit in CNS environment- if heketi does the check, or has NO way of going wrong with the IPs. Will need to confirm on this. Version-Release number of selected component (if applicable): ============================================================ glusterfs-3.8.4-33 and gluster-block-0.2.1-6 How reproducible: ================ 3:3 Steps to Reproduce: =================== 1. Create a 1*3 volume, on a 3node cluster setup 2. Create a block of 'ha=3' and give addresses of node1, node2 and again node1. Actual results: ============== Command exits, after a long time with Broken pipe. Login to the console shows no control on the cli. Force reboot the VM, and it is back up. Expected results: ================ Block create command should check the validity of addresses, before proceeding with the actual block creation. Additional info: ================ Did not get much information from dmesg or /var/log/messages. [root@dhcp47-116 abrt]# gluster-block create nash/nb55 ha 3 10.70.47.116,10.70.47.117,10.70.47.116 1M packet_write_wait: Connection to 10.70.47.116 port 22: Broken pipe bash-4.3$ ssh root.47.116 ^C bash-4.3$ [root@dhcp47-115 ~]# cat /mnt/nash/block-meta/nb55 VOLUME: nash GBID: 5115979d-9a8d-4ea8-bcae-23280e9c6e2f SIZE: 1048576 HA: 3 ENTRYCREATE: INPROGRESS ENTRYCREATE: SUCCESS 10.70.47.116: CONFIGINPROGRESS 10.70.47.116: CONFIGINPROGRESS 10.70.47.117: CONFIGINPROGRESS 10.70.47.116: CONFIGSUCCESS 10.70.47.117: CONFIGSUCCESS 10.70.47.116: CONFIGFAIL 10.70.47.117: CLEANUPINPROGRESS [root@dhcp47-115 ~]# [root@dhcp47-115 ~]# rpm -qa | grep gluster glusterfs-cli-3.8.4-33.el7rhgs.x86_64 glusterfs-rdma-3.8.4-33.el7rhgs.x86_64 python-gluster-3.8.4-33.el7rhgs.noarch vdsm-gluster-4.17.33-1.1.el7rhgs.noarch glusterfs-client-xlators-3.8.4-33.el7rhgs.x86_64 glusterfs-fuse-3.8.4-33.el7rhgs.x86_64 gluster-nagios-common-0.2.4-1.el7rhgs.noarch glusterfs-events-3.8.4-33.el7rhgs.x86_64 gluster-block-0.2.1-6.el7rhgs.x86_64 libvirt-daemon-driver-storage-gluster-3.2.0-14.el7.x86_64 gluster-nagios-addons-0.2.9-1.el7rhgs.x86_64 samba-vfs-glusterfs-4.6.3-3.el7rhgs.x86_64 glusterfs-3.8.4-33.el7rhgs.x86_64 glusterfs-debuginfo-3.8.4-26.el7rhgs.x86_64 glusterfs-api-3.8.4-33.el7rhgs.x86_64 glusterfs-geo-replication-3.8.4-33.el7rhgs.x86_64 glusterfs-libs-3.8.4-33.el7rhgs.x86_64 glusterfs-server-3.8.4-33.el7rhgs.x86_64 [root@dhcp47-115 ~]# [root@dhcp47-115 ~]# gluster peer status Number of Peers: 5 Hostname: dhcp47-121.lab.eng.blr.redhat.com Uuid: 49610061-1788-4cbc-9205-0e59fe91d842 State: Peer in Cluster (Connected) Other names: 10.70.47.121 Hostname: dhcp47-113.lab.eng.blr.redhat.com Uuid: a0557927-4e5e-4ff7-8dce-94873f867707 State: Peer in Cluster (Connected) Hostname: dhcp47-114.lab.eng.blr.redhat.com Uuid: c0dac197-5a4d-4db7-b709-dbf8b8eb0896 State: Peer in Cluster (Connected) Other names: 10.70.47.114 Hostname: dhcp47-116.lab.eng.blr.redhat.com Uuid: a96e0244-b5ce-4518-895c-8eb453c71ded State: Peer in Cluster (Disconnected) Other names: 10.70.47.116 Hostname: dhcp47-117.lab.eng.blr.redhat.com Uuid: 17eb3cef-17e7-4249-954b-fc19ec608304 State: Peer in Cluster (Connected) Other names: 10.70.47.117 [root@dhcp47-115 ~]# [root@dhcp47-115 ~]# [root@dhcp47-115 ~]# gluster v info nash Volume Name: nash Type: Replicate Volume ID: f1ea3d3e-c536-4f36-b61f-cb9761b8a0a6 Status: Started Snapshot Count: 0 Number of Bricks: 1 x 3 = 3 Transport-type: tcp Bricks: Brick1: 10.70.47.115:/bricks/brick4/nash0 Brick2: 10.70.47.116:/bricks/brick4/nash1 Brick3: 10.70.47.117:/bricks/brick4/nash2 Options Reconfigured: nfs.disable: on transport.address-family: inet performance.quick-read: off performance.read-ahead: off performance.io-cache: off performance.stat-prefetch: off performance.open-behind: off performance.readdir-ahead: off network.remote-dio: enable cluster.eager-lock: enable cluster.quorum-type: auto cluster.data-self-heal-algorithm: full cluster.locking-scheme: granular cluster.shd-max-threads: 8 cluster.shd-wait-qlength: 10000 features.shard: on user.cifs: off server.allow-insecure: on cluster.brick-multiplex: disable cluster.enable-shared-storage: enable [root@dhcp47-115 ~]# [root@dhcp47-115 ~]# [root@dhcp47-115 ~]#
Not proposing this as a blocker for rhgs3.3 as I have got an input from CNS QE that heketi will not execute 'gluster-block create' command with duplicate IPs.
Moving it to 'UPSTREAM' as this is not the usecase, and the tools creating block files are not calling the CLI like this.