Comment 11 for bug 1722778

Since this bug appeared when I upgraded from zesty to artful, I reverted to 4.10.0-37 (the last kernel I had in zesty) and it had been rock-solid (it becomes unresponsive from time to time but it recovers in a few seconds).
With each kernel upgrade I try the new one but the problem is still there (though I did not try 4.13.0-31, but I don't see anything in its changelog regarding this bug).
I also tried the BFQ scheduler, (as suggested to me here http://ubuntu.5.x6.nabble.com/Bad-memory-swap-management-in-17-10-td5160226.html) but it didn't resolve the issue.
Regarding the previous comment than "4 GBs ram is quite small", 4.10.0 manages quite well the same workload (I have 8Gb but four are eaten by virtualbox).