He probably didn't do anything wrong. Deleting log files which have no function serves no purpose. There is an underlying problem. Restarting, rescanning and repairing seems to fix temporarily but the error always comes back. Sometimes after a few minutes sometimes after hours, sometimes after a re-start of the client. It's not the size of the log file that is a problem. If you start the client with full logging enabled the log file ends up ten times the size. It doesn't seem to stop the wallet working though. I've been getting stakes too, whilst the error is displayed but often when re-starting the wallet I end up with double my real balance until a further restart/rescan.........not too worried. It'll get resolved.
What I see at the debug.log file is:
Added time data, samples 14, offset -3 (+0 minutes)
version message: version 60008, blocks=361988
REORGANIZE
REORGANIZE: Disconnect 10294 blocks; c5d58fb33aaef0f16a6c..6ee7b223b9ce545c0d39
REORGANIZE: Connect 56 blocks; c5d58fb33aaef0f16a6c..4c2a420566853c9a6282
************************
EXCEPTION: 11DbException
Db::put: Cannot allocate memory
Snowballs in ProcessMessages()
ProcessMessage(checkpoint, 110 bytes) FAILED
I suspect that a node running an old version 2.0.1 tries to connect and causes a reorganization, but fails. Just after the fork the 2.0.1 nodes were successful because the blocks to reorganize were only a few, but now 10,000 blocks are too many blocks.
I've been able to stake today after getting that message, so I guess it is not a big issue ...