Comment 1 for bug 1595940

This job ran on jenkins-slave-0 when it had problems with slow CPU and I/O performance (note that the linked job took 1h22m to finish, the normal time being ~40m).

My educated guess is that this failure was caused by the extreme slowness, which led to the test exceeding the relatively short (1s) timeout used to wait for 'enough_records_gathered'.

Marking as incomplete for now (so it can expire). If we see this again, we can consider increasing the timeout.