Kadalu Storage Volume Rebalance #26
Conversation
aravindavk
force-pushed
the
rebalance
branch
from
94209e2 to
adb12e0
Compare
| kadalu volume rebalance-stop <POOL>/<VOLUME> | ||
| ---- | ||
|
|
||
| Above command calls the respective API that internally starts a service in respective Storage unit nodes. **Note**: These services will halt once the Rebalance process completes its job. Do not start the service in every nodes of the Volume, start only in the node of the first Storage unit from each distribute groups. |
There was a problem hiding this comment.
The design is cool 💯
Any reason for the command to be run from node of 1st storage unit from each distribute group. Is it avoid redundancy or wrong calculation of starting point of hash range.
Will it not be feasible to call the command from server node as an option, since moana has server-client model and create a node action to fetch necessary xattr details from nodes of 1st storage unit of dist grp to make it much simpler?
There was a problem hiding this comment.
First node is to make it predictable. Any random node is fine, but Moana should keep a reference where it started the service. We can improve this part later. For now it always starts in the first node of every distribute group.
| kubectl kadalu rebalance-status <Pool-name> | ||
| ---- | ||
|
|
||
| == Tool to handle the rebalance |
There was a problem hiding this comment.
Does this also contain the migrate data option to remove linkfiles created for renamed files in cached locations or this is part of glusterfs rebalance command itself?
There was a problem hiding this comment.
Migrate data is handled by GlusterFS itself via the exposed virtual xattrs. Rebalance tool only manages the filesystem crawl and calls Rebalance xattrs.
Signed-off-by: Aravinda Vishwanathapura <mail@aravindavk.in>
Signed-off-by: Aravinda Vishwanathapura mail@aravindavk.in