Hi, One of the jobs running on our cluster died. After investigating logs I can clearly see that job died because client that was running that job was evicted by MDS Nov 14 13:31:27 node-i06 kernel: LustreError: 167-0: This client was evicted by ddn_home-MDT0000; in progress operations using this service will fail. On MDS I can not see any messages about this client being evicted. There is only one message about this client: Nov 14 13:31:27 mds01.beowulf.cluster kernel: LustreError: 22512:0: (handler.c:1498:mds_handle()) operation 400 on unconnected MDS from 12345-10.143.9.6 at tcp Can some one explain me what exactly had happened? Is this problem can be related to https://bugzilla.lustre.org/show_bug.cgi?id=13682 Best regards, Wojciech Turek -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.lustre.org/pipermail/lustre-discuss/attachments/20071114/845ee0ef/attachment-0002.html