We have seen something like this on our HP DL380 Servers. This usually appeared when the Disk subsystem was not able to keep up with the data being pumped out by the processor. Turn out the HW raid available for HP proliants(Smart Array) has very poor performance at best( 20-30MB/s throughput). We went to software raid, and the problem was greatly diminished, as we get arount 100MB/s performance, and these " in journal_alloc_journal_head" errors pretty much went away. Our server was: dl380 2x2.8GHz Xeon, 6GB RAM, RedHAt Advanced Server2.1, Oracle 9i. Dan -----Original Message----- From: Anis Widyanto Nugroho [mailto:anis_wn@cbn.net.id] Sent: Thursday, July 31, 2003 4:57 PM To: ext3-users@redhat.com Subject: ENOMEM in journal_ Sorry for incomplete message... Hi all, We have a problem with our system which is a Siemens PRIMERGY, 8xPentium III Xeon, 8GB Memory, 280GB HD, running as DB2 server. We have some kind of error message like below in /var/log/messages: .. Jul 31 10:56:42 dbserver kernel: ENOMEM in journal_alloc_journal_head, retrying. Jul 31 10:57:54 dbserver kernel: ENOMEM in journal_get_undo_access_Rsmp_df5dec49, retrying. .. Jul 31 14:16:07 dbserver kernel: ENOMEM in journal_alloc_journal_head, retrying. Jul 31 14:16:52 dbserver kernel: ENOMEM in do_get_write_access, retrying. Jul 31 14:16:58 dbserver kernel: ENOMEM in journal_alloc_journal_head, retrying. Jul 31 14:17:33 dbserver kernel: ENOMEM in journal_alloc_journal_head, retrying. .. All the case, lead our machine to hang. Any suggestion about settings on our machine? Perhaps on sysctl, etc. Thanks in advance. Regards, Anis WN _______________________________________________ Ext3-users@redhat.com https://www.redhat.com/mailman/listinfo/ext3-users _______________________________________________ Ext3-users@redhat.com https://www.redhat.com/mailman/listinfo/ext3-users
