Hitoshi Mitake
Hitoshi Mitake
Fix https://github.com/etcd-io/etcd/issues/12385 It's still a WIP PR, please do not merge. Remaining todos: - [ ] clean up code - [ ] write a test
… snapshot Fix https://github.com/etcd-io/etcd/issues/13883 The above issue reports a problem that `authStore.Recover()` doesn't invalidate `rangePermCache`, so an etcd node which is isolated from its cluster might not invalidate stale permission...
We need to describe what the default option --unsafe implies in case of node disconnection and what --strict does.
Current sheepdog support runtime monitoring with "dog node stat". Although it is useful, metrics and interface should be enhanced e.g. plotting graph of resource usage will be useful. http://prometheus.io/ seems...
0 192.168.2.44:7000 3 738371776 1 192.168.2.45:7000 3 755148992 2 192.168.2.46:7000 3 771926208 3 192.168.2.47:7000 3 788703424 1) I create a test disk dog vdi create -P test 10G 2) I...
journaling is very unstable component and seems no to contribute to performance
Quoting email from @imrook My understanding from the page at https://github.com/sheepdog/sheepdog/wiki/Which-Format-of-QEMU-Images-Should-I-Run is that sheepdog can store qemu checkpoints internally using the 'raw' format. I am having problems when trying to...
Current sheep can call `leave_cluster()` when it faces EIO, e.g. https://github.com/sheepdog/sheepdog/blob/master/sheep/request.c#L123 This is clearly nonsense. In a case of single drive node, the broken disk drive would be shared with...
Clearly 6,144 nodes are out of sheepdog's scope. Reducing the maximum nodes from 6144 to 1024 will simplify the implementation because every cluster driver can have the same maximum number...
On a single host with no data /usr/sbin/sheep -n /var/lib/sheepdog,/mnt/sheep/0,/mnt/sheep/1,/mnt/sheep/2 --cluster zookeeper:192.168.6.45:2181 cat /var/lib/sheepdog/sheep.log Aug 19 16:25:28 INFO [main] md_add_disk(368) /mnt/sheep/0, vdisk nr 413, total disk 1 Aug 19 16:25:28...