[lustre-discuss] issue: strange pauses between writes, but not everywhere

[Peter Grandi]

>>> On Wed, 29 Oct 2025 12:45:52 +0000, Peter Grandi via
>>> lustre-discuss <lustre-discuss at lists.lustre.org> said:

>> * [...] Reading is quite good for all clients. But since
>> yesterday early afternoon inexplicably the clients .13-36
>> have a maximum average write speed of around 35-40MB/s [...]

So I tried to run 3 long writes on the same clients to the same
Lustre instance 'temp01' and directory and each of the writes
was running at that speed (actually a bit better, around 52MB/s
each), plus a write to the the same instance via NFSv4 and that
was running at around 380MB/s.

So it is per-client-process (or per-client-connection) not quite
per-client-host which seems to confirm that it is not a straight
networking issue that suddenly emerged (again clients and
servers in these tests are connected to the same switch).