You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi swiftgist,
We've had a critical issue when try use multi gateways, multipath active/active and enable "round-robin" policy in client configuration. With this configuration, the data will be inconsistent because the data maybe in cache not in Ceph cluster already for the other GW access. You should add this limitation to your wiki. "active/standby" is recommendation and with this config, it will limit just one tpg on one gateway node.
To solve this issue, tried to config alua feature. I seem to success to help me control the paths of all clients connect to my gateways. Right now for one LUN, the data traffic just come and out at same gateway node.
The text was updated successfully, but these errors were encountered:
Hi swiftgist,
We've had a critical issue when try use multi gateways, multipath active/active and enable "round-robin" policy in client configuration. With this configuration, the data will be inconsistent because the data maybe in cache not in Ceph cluster already for the other GW access. You should add this limitation to your wiki. "active/standby" is recommendation and with this config, it will limit just one tpg on one gateway node.
To solve this issue, tried to config alua feature. I seem to success to help me control the paths of all clients connect to my gateways. Right now for one LUN, the data traffic just come and out at same gateway node.
The text was updated successfully, but these errors were encountered: