[ClusterLabs] questions about fence_scsi

[Andrei Borzenkov]

On Fri, Jun 15, 2018 at 11:18 AM, Andrei Borzenkov <arvidjaar at gmail.com> wrote:
> On Fri, Jun 15, 2018 at 10:14 AM, Stefan Krueger <Shadow_7 at gmx.net> wrote:
>> Hello,
>>
>> so far as I understand I can use fence_scsi on a two node cluster, if the fence running on one cluster the other cluster has no access to this devices, correct?
>
> If I parse this sentence correctly - no, that's not correct to my best
> knowledge. All active nodes have access to shared resource - only when
> node fails will it be fenced (i.e. access to devices revoked), and
> unfenced again (i.e. access granted) when node comes back.
>

If you really want to allow only one node at a time to access device,
you should look at something like sg_persist RA. See example in SLES
focumentation (scroll down to the end of page):

https://www.suse.com/documentation/sle-ha-12/book_sleha/data/sec_ha_storage_protect_fencing.html


>> I've a 2node cluster with shared JBODs and configure fence_scsi, but I still can use/mount all this devices on both nodes. Did I something wrong?
>> My Config:
>>
>> pcs resource
>>  Resource Group: zfs-storage
>>      vm_storage (ocf::heartbeat:ZFS):   Started zfs-serv3
>>      ha-ip      (ocf::heartbeat:IPaddr2):       Started zfs-serv3
>> root at zfs-serv4:~# pcs config
>> Cluster Name: zfs-vmstorage
>> Corosync Nodes:
>>  zfs-serv3 zfs-serv4
>> Pacemaker Nodes:
>>  zfs-serv3 zfs-serv4
>>
>> Resources:
>>  Group: zfs-storage
>>   Resource: vm_storage (class=ocf provider=heartbeat type=ZFS)
>>    Attributes: pool=vm_storage importargs="-d /dev/disk/by-vdev/"
>>    Operations: monitor interval=5s timeout=30s (vm_storage-monitor-interval-5s)
>>                start interval=0s timeout=90 (vm_storage-start-interval-0s)
>>                stop interval=0s timeout=90 (vm_storage-stop-interval-0s)
>>   Resource: ha-ip (class=ocf provider=heartbeat type=IPaddr2)
>>    Attributes: ip=172.16.101.73 cidr_netmask=16
>>    Operations: start interval=0s timeout=20s (ha-ip-start-interval-0s)
>>                stop interval=0s timeout=20s (ha-ip-stop-interval-0s)
>>                monitor interval=10s timeout=20s (ha-ip-monitor-interval-10s)
>>
>> Stonith Devices:
>>  Resource: fence-vm_storage (class=stonith type=fence_scsi)
>>   Attributes: pcmk_monitor_action=metadata pcmk_host_list=172.16.101.74,172.16.101.75 devices=" /dev/disk/by-vdev/j3d03-hdd /dev/disk/by-vdev/j4d03-hdd /dev/disk/by-vdev/j3d04-hdd /dev/disk/by-vdev/j4d04-hdd /dev/disk/by-vdev/j3d05-hdd /dev/disk/by-vdev/j4d05-hdd /dev/disk/by-vdev/j3d06-hdd /dev/disk/by-vdev/j4d06-hdd /dev/disk/by-vdev/j3d07-hdd /dev/disk/by-vdev/j4d07-hdd /dev/disk/by-vdev/j3d08-hdd /dev/disk/by-vdev/j4d08-hdd /dev/disk/by-vdev/j3d09-hdd /dev/disk/by-vdev/j4d09-hdd /dev/disk/by-vdev/j3d10-hdd /dev/disk/by-vdev/j4d10-hdd /dev/disk/by-vdev/j3d11-hdd /dev/disk/by-vdev/j4d11-hdd /dev/disk/by-vdev/j3d12-hdd /dev/disk/by-vdev/j4d12-hdd /dev/disk/by-vdev/j3d13-hdd /dev/disk/by-vdev/j4d13-hdd /dev/disk/by-vdev/j3d14-hdd /dev/disk/by-vdev/j4d14-hdd /dev/disk/by-vdev/j3d15-hdd /dev/disk/by-vdev/j4d15-hdd /dev/disk/by-vdev/j3d16-hdd /dev/disk/by-vdev/j4d16-hdd /dev/disk/by-vdev/j3d17-hdd /dev/disk/by-vdev/j4d17-hdd /dev/disk/by-vdev/j3d18-hdd /dev/disk/by-vdev/j4d18-hdd /dev/d
>>  isk/by-vdev/j3d19-hdd /dev/disk/by-vdev/j4d19-hdd log /dev/disk/by-vdev/j3d00-ssd /dev/disk/by-vdev/j4d00-ssd cache /dev/disk/by-vdev/j3d02-ssd"
>>   Meta Attrs: provides=unfencing
>>   Operations: monitor interval=60s (fence-vm_storage-monitor-interval-60s)
>> Fencing Levels:
>>
>> Location Constraints:
>> Ordering Constraints:
>> Colocation Constraints:
>> Ticket Constraints:
>>
>> Alerts:
>>  No alerts defined
>>
>> Resources Defaults:
>>  resource-stickiness: 100
>> Operations Defaults:
>>  No defaults set
>>
>> Cluster Properties:
>>  cluster-infrastructure: corosync
>>  cluster-name: zfs-vmstorage
>>  dc-version: 1.1.16-94ff4df
>>  have-watchdog: false
>>  last-lrm-refresh: 1528814481
>>  no-quorum-policy: ignore
>>
>> Quorum:
>>   Options:
>>
>>
>> Thanks for help!
>> Best regards
>> Stefan
>> _______________________________________________
>> Users mailing list: Users at clusterlabs.org
>> https://lists.clusterlabs.org/mailman/listinfo/users
>>
>> Project Home: http://www.clusterlabs.org
>> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
>> Bugs: http://bugs.clusterlabs.org