<html><body><div style="font-family: garamond, new york, times, serif; font-size: 12pt; color: #000000"><div data-marker="__QUOTED_TEXT__"><div style="font-family:'garamond' , 'new york' , 'times' , serif;font-size:12pt;color:#000000"><div>Hello,<br></div><br><div>I am new on this forum.<br></div><br><div>We operate a Lustre filesystem (Lustre 2.12.16 on CentOS 7.9 servers, <span class="search_hit">lustre</span>-client-2.12.4-1.el7.x86_64.rpm and <span class="search_hit">lustre</span>-client-dkms-2.12.4-1.el7.noarch.rpm on CentOS 7.3 clients, simple installation following official documentation on wiki Lustre).<br></div><br><div>When data processing is consuming about 75%-80% of CPU cores on a client :<br></div><div>- lustre performance decrease but only on the overload client (normal performance on other clients)<br></div><div>- system performance decrease on the overload client, with %CPU sys much higher than similar systems working on local filesystems<br></div><div>- nothing to notice on lustre servers<br></div><br><div>This happens even when attending short peaks of CPU consumption and whatever the data processing.<br></div><br><div>Have you ever notice such troubles on your installations ?<br></div><div>If yes, is there any solution to this problem ?<br></div><br><div>Best regards,<br></div><br><div>Karine<br></div></div><br></div></div></body></html>