Block storage performance has been terrible for the last week
Since around mindnight (CET) between 4. and 5. of November, access to files stored on block storage has been terrible. Especially accessing lots of files in a short timespan (e.g running a backup etc) is sometimes taking much, much longer than it used to do.
I can see processes that normally finish in 3-4 minutes now taking almost an hour to complete, and the droplet is almost unusable while it is running, due to IO-Wait which again leads to other processes timing out, and extreme high memory usage.
I can clearly see this in the IO, Network, load avarega and CPU and Memory graphs, where there is a clear correlation bewteen the different graphs, and when this happens. The Storage-graph on the other hand, does not show any significant change in usage before or after November 5. So I do not believe this is due to something I have done on the server.
Did anything change with block storage in Amsterdam at that time?
Is there anything I can do to mitigate this problem?