Whamcloud Community JIRA https://jira.whamcloud.com This file is an XML representation of an issue en-us 9.4.14 940014 05-12-2023 [LU-4709] Can't allocate super-sequence, rc -115 https://jira.whamcloud.com/browse/LU-4709 Lustre <p>In setting up a Lustre 2.5.0 environment I am encountering an error.</p> <div class="preformatted panel" style="border-width: 1px;"><div class="preformattedContent panelContent"> <pre>&lt;normal lustre mount messages&gt; 23:35:56 lotus-24 kernel: LDISKFS-fs (sdk8): mounted filesystem with ordered data mode. quota=on. Opts: Mar 3 23:35:56 lotus-24 kernel: LDISKFS-fs (sdk8): mounted filesystem with ordered data mode. quota=on. Opts: Mar 3 23:35:57 lotus-24 kernel: LDISKFS-fs (sdk8): mounted filesystem with ordered data mode. quota=on. Opts: Mar 3 23:42:18 lotus-24 kernel: LustreError: 17341:0:(fid_handler.c:286:__seq_server_alloc_meta()) srv-testfs-OST000f: Can't allocate super-sequence, rc -115 Mar 3 23:42:18 lotus-24 kernel: LustreError: 17341:0:(fid_handler.c:286:__seq_server_alloc_meta()) Skipped 1392 previous similar messages Mar 3 23:52:18 lotus-24 kernel: LustreError: 15693:0:(fid_handler.c:286:__seq_server_alloc_meta()) srv-testfs-OST0002: Can't allocate super-sequence, rc -115 Mar 3 23:52:18 lotus-24 kernel: LustreError: 15693:0:(fid_handler.c:286:__seq_server_alloc_meta()) Skipped 1477 previous similar messages Mar 4 00:02:18 lotus-24 kernel: LustreError: 17341:0:(fid_handler.c:286:__seq_server_alloc_meta()) srv-testfs-OST001b: Can't allocate super-sequence, rc -115 Mar 4 00:02:18 lotus-24 kernel: LustreError: 17341:0:(fid_handler.c:286:__seq_server_alloc_meta()) Skipped 1466 previous similar messages Mar 4 00:12:19 lotus-24 kernel: LustreError: 15693:0:(fid_handler.c:286:__seq_server_alloc_meta()) srv-testfs-OST0009: Can't allocate super-sequence, rc -115 Mar 4 00:12:19 lotus-24 kernel: LustreError: 15693:0:(fid_handler.c:286:__seq_server_alloc_meta()) Skipped 1477 previous similar messages Mar 4 00:22:20 lotus-24 kernel: LustreError: 17341:0:(fid_handler.c:286:__seq_server_alloc_meta()) srv-testfs-OST0018: Can't allocate super-sequence, rc -115 ..... 08:52:34 lotus-24 kernel: LustreError: 17341:0:(fid_handler.c:286:__seq_server_alloc_meta()) srv-testfs-OST0004: Can't allocate super-sequence, rc -115 Mar 4 08:52:34 lotus-24 kernel: LustreError: 17341:0:(fid_handler.c:286:__seq_server_alloc_meta()) Skipped 1474 previous similar messages Mar 4 09:02:34 lotus-24 kernel: LustreError: 22775:0:(fid_handler.c:286:__seq_server_alloc_meta()) srv-testfs-OST000f: Can't allocate super-sequence, rc -115 Mar 4 09:02:34 lotus-24 kernel: LustreError: 22775:0:(fid_handler.c:286:__seq_server_alloc_meta()) Skipped 1470 previous similar messages Mar 4 09:12:34 lotus-24 kernel: LustreError: 17341:0:(fid_handler.c:286:__seq_server_alloc_meta()) srv-testfs-OST0002: Can't allocate super-sequence, rc -115 Mar 4 09:12:34 lotus-24 kernel: LustreError: 17341:0:(fid_handler.c:286:__seq_server_alloc_meta()) Skipped 1477 previous similar messages Mar 4 09:22:34 lotus-24 kernel: LustreError: 15693:0:(fid_handler.c:286:__seq_server_alloc_meta()) srv-testfs-OST001b: Can't allocate super-sequence, rc -115 Mar 4 09:22:34 lotus-24 kernel: LustreError: 15693:0:(fid_handler.c:286:__seq_server_alloc_meta()) Skipped 1466 previous similar messages Mar 4 09:32:35 lotus-24 kernel: LustreError: 20339:0:(fid_handler.c:286:__seq_server_alloc_meta()) srv-testfs-OST0009: Can't allocate super-sequence, rc -115 Mar 4 09:32:35 lotus-24 kernel: LustreError: 20339:0:(fid_handler.c:286:__seq_server_alloc_meta()) Skipped 1477 previous similar messages </pre> </div></div> <p>df shows all the OSTs mounted and present but trying to mount the namespace on a Client seems to reboot the systems. </p> <p>As present there are 27 or so OSTs and a MDT on this server, this same behavior is seen on pure OSSs. The Lustre targets are physically slow. </p> Lustre 2.5.0 Development environment, system is fresh and /scratch LU-4709 Can't allocate super-sequence, rc -115 Bug Minor Resolved Not a Bug WC Triage Keith Mannthey Tue, 4 Mar 2014 17:58:16 +0000 Wed, 5 Mar 2014 01:34:47 +0000 Wed, 5 Mar 2014 01:34:47 +0000 0 1 <p>It can also been seen in mount blocks like:</p> <div class="preformatted panel" style="border-width: 1px;"><div class="preformattedContent panelContent"> <pre>LDISKFS-fs (sdk3): mounted filesystem with ordered data mode. quota=on. Opts: LDISKFS-fs (sdk3): mounted filesystem with ordered data mode. quota=on. Opts: LDISKFS-fs (sdk3): mounted filesystem with ordered data mode. quota=on. Opts: Lustre: testfs-OST0011: new disk, initializing Lustre: Skipped 8 previous similar messages LustreError: 13a-8: Failed to get MGS log params and no local copy. LustreError: Skipped 17 previous similar messages LustreError: 44563:0:(fid_handler.c:286:__seq_server_alloc_meta()) srv-testfs-OST0001: Can't allocate super-sequence, rc -115 LustreError: 44563:0:(fid_handler.c:286:__seq_server_alloc_meta()) Skipped 340 previous similar messages LDISKFS-fs (sdk5): mounted </pre> </div></div> <p>Well it seems this is what happens when you create a -mdt -mgt with index=1 instead of index=0 (when you only have a single MDT). User error off-by-one. </p> <p>Something less then a client crash would be nice but it was user error. </p> Development Rank 1|hzwgq7: Rank (Obsolete) 12946 Severity