On Tue, Jun 22, 2021 at 2:01, Dominique D
<dominique.deschenes@gcgenicom.com> wrote:
yesterday I had a disk failure on my stack of 3 Ovirt 4.4.1 node
on each server I have 3 Bricks (engine, data, vmstore)
brick data 4X600Gig raid0. /dev/gluster_vg_sdb/gluster_lv_data mount /gluster_bricks/data
brick engine 2X1TB raid1 /dev/gluster_vg_sdc/gluster_lv_engine mount /gluster_bricks/engine
brick vmstore 2X1TB raid1 /dev/gluster_vg_sdc/gluster_lv_vmstore mount /gluster_bricks/vmstore
Everything was configured by the gui (hyperconverge and hosted-engine)
It is the raid0 of the 2nd server who broke.
all VMs were automatically moved to the other two servers, I haven't lost any data.
the host2 is now in maintenance mode.
I am going to buy 4 new SSD disks to replace the 4 disks of the defective raid0.
When I'm going to erase the faulty raid0 and create the new raid with the new disks on the raid controler, how do I add in ovirt so that they resynchronize with the other bricks data?
Status of volume: data
Gluster process TCP Port RDMA Port Online Pid
------------------------------------------------------------------------------
Brick 172.16.70.91:/gluster_bricks/data/dat
a 49153 0 Y 79168
Brick 172.16.70.92:/gluster_bricks/data/dat
a N/A N/A N N/A
Brick 172.16.70.93:/gluster_bricks/data/dat
a 49152 0 Y 3095
Self-heal Daemon on localhost N/A N/A Y 2528
Self-heal Daemon on 172.16.70.91 N/A N/A Y 225523
Self-heal Daemon on 172.16.70.93 N/A N/A Y 3121
_______________________________________________