Hi, We got through the errors we had earlier this week with hard locking when mounting specific OSTs. The same OST (ost7) that was causing locks now mounts but cannot be communicated with it turns out. I''m getting hundreds of the following errors in my logs: Oct3 13:32:52 iroc kernel: Lustre Error:5400:0:(filter.c261:filter_client_add()) no room for 32768 clients - fix LR_MAX_CLIENTS Oct3 13:36:17 iroc kernel: LustreErro:5364:0:(fsfilt-ldiskfs.c:282:fsfilt_ldiskfs_start()) error starting handle for op 4 (51 credits):rc -30 Oct3 13:36:19 iroc kernel:LustreErro:8756:0:(fsfilt-ldiskfs.c:282:fsfilt_ldiskfs_start()) error starting handle for op 4 (51 credits):rc -30 Oct3 13:36:17 iroc kernel:LustreErro:5363:0:(fsfilt-ldiskfs.c:282:fsfilt_ldiskfs_start()) error starting handle for op 4 (51 credits):rc -30 Oct3 13:37:57 iroc kernel: LustreError: 5582:0:(ldlm_lib.c:1536:target_send_reply_msg()) @@@ processing error (-75) req at 000010120c70e00x4470/t0 o8-><?>:0/0 lens 240/144 e 0 to 0 dl 1223066377 ref 1 fl Interpret:/0/0 rc -75/0 What do these errors indicate? Clients are reporting timeouts trying to reach ost7 and failing. This locks the file system for minutes at a time. Thanks, Dan
On Fri, 2008-10-03 at 13:57 -0700, Dan wrote:> Hi, > > We got through the errors we had earlier this week with hard locking when mounting specific OSTs. The same OST (ost7) that was causing locks now mounts but cannot be communicated with it turns out. > > I''m getting hundreds of the following errors in my logs: > > > Oct3 13:32:52 iroc kernel: Lustre > Error:5400:0:(filter.c261:filter_client_add()) no room for 32768 clients > - fix LR_MAX_CLIENTSThis looks like bug 17052.> Oct3 13:36:17 iroc kernel: > LustreErro:5364:0:(fsfilt-ldiskfs.c:282:fsfilt_ldiskfs_start()) error starting handle for op 4 (51 credits):rc -30 > Oct3 13:36:19 iroc > kernel:LustreErro:8756:0:(fsfilt-ldiskfs.c:282:fsfilt_ldiskfs_start()) error starting handle for op 4 (51 credits):rc -30 > Oct3 13:36:17 iroc > kernel:LustreErro:5363:0:(fsfilt-ldiskfs.c:282:fsfilt_ldiskfs_start()) error starting handle for op 4 (51 credits):rc -30-30 is EROFS, which means that the backing store device was set to read-only. You need to search through your logs for why. b. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 197 bytes Desc: This is a digitally signed message part Url : http://lists.lustre.org/pipermail/lustre-discuss/attachments/20081006/8e8e1b18/attachment.bin
Hi, Can you expand on bug 17052 - I don''t have access to it in bugzilla. I''ll check for anything mounted ro... I didn''t think it was! Thank you! Dan Brian J. Murrell wrote: On Fri, 2008-10-03 at 13:57 -0700, Dan wrote: Hi, We got through the errors we had earlier this week with hard locking when mounting specific OSTs. The same OST (ost7) that was causing locks now mounts but cannot be communicated with it turns out. I''m getting hundreds of the following errors in my logs: Oct3 13:32:52 iroc kernel: Lustre Error:5400:0:(filter.c261:filter_client_add()) no room for 32768 clients - fix LR_MAX_CLIENTS This looks like bug 17052. Oct3 13:36:17 iroc kernel: LustreErro:5364:0:(fsfilt-ldiskfs.c:282:fsfilt_ldiskfs_start()) error starting handle for op 4 (51 credits):rc -30 Oct3 13:36:19 iroc kernel:LustreErro:8756:0:(fsfilt-ldiskfs.c:282:fsfilt_ldiskfs_start()) error starting handle for op 4 (51 credits):rc -30 Oct3 13:36:17 iroc kernel:LustreErro:5363:0:(fsfilt-ldiskfs.c:282:fsfilt_ldiskfs_start()) error starting handle for op 4 (51 credits):rc -30 -30 is EROFS, which means that the backing store device was set to read-only. You need to search through your logs for why. b. _______________________________________________ Lustre-discuss mailing list Lustre-discuss-aLEFhgZF4x6X6Mz3xDxJMA@public.gmane.org http://lists.lustre.org/mailman/listinfo/lustre-discuss _______________________________________________ Lustre-discuss mailing list Lustre-discuss-aLEFhgZF4x6X6Mz3xDxJMA@public.gmane.org http://lists.lustre.org/mailman/listinfo/lustre-discuss