Sorry. Unfortunately, today (on Friday) we can't run the round 383 :-(
In short, this is because of technical reasons. For those interested in a more detailed description option.
On Thursday night I got sms. It was written that on server for DB backups free HDD space is less than 10%. This server stores three copies of the database over the last three days. It has 2TB drive and Codeforces database over 600GB now.
Since the next day I had planned to spend on the train on the way to NEERC, I decided to delete some redundant data from the DB. I estimate it will give several tens of gigabytes of free space.
I run query to remove redundant data and it was a mistake. I seems because of huge table to update (tens millions of rows) estimated time for the query to compete was at least several hours. As a result, I have killed this query, making KILL QUERY. But the database has continued to lag, I decided to restart it. Operation to shutdown the DB hanged for ~2 hours (I think it was rolling back the transaction), and the DB has stopped to respond and has become quite sad. As a result, I killed the DB process in the operating system and the database already refused to start without a recovery operation, which is also wasn't completed in a couple of hours. In short, I have struggled with this from 5 am to 9 am.
To make Codeforces live I reconfigured it to use database replica in the remote datacenter (in Mail.Ru). Before boarding the train, I started the process of transferring data from the replica back, but contrary to my expectations, this process has been very slow.
It is impossible to run the round using a remote reserve database server. Therefore, we have to move the round forward to Tuesday.
My apologies. I spent many hours struggling with technical issues and worked up to the last. However, it seems the bottleneck is in speed of data transmission over a network, in the speed to write on HDD and in the speed of innobackupex utility.
I hope that on Tuesday everything will go fine!