Due to the expansion of our cluster, disk i/o performance is degraded. We will restore the usual performance as soon as possible.

Update 12:52

The expansion is complete, and latency is back to normal. Due to the added hardware, we do not expect any further incidents.