Prevent nodes from reconnecting too fast
This happened between storage nodes of different clusters because they're not informed about their state, e.g. a dead upstream storage node. In any case, logs were flooded at 100% cpu usage.
Showing
Please register or sign in to comment