Bug #74829 | ndb.ndb_restore_conv_more fails on POWER8 | ||
---|---|---|---|
Submitted: | 13 Nov 2014 3:47 | Modified: | 18 Jan 2016 20:59 |
Reporter: | Stewart Smith | Email Updates: | |
Status: | No Feedback | Impact on me: | |
Category: | MySQL Cluster: Cluster (NDB) storage engine | Severity: | S2 (Serious) |
Version: | 7.3.7 | OS: | Linux |
Assigned to: | Assigned Account | CPU Architecture: | Any |
Tags: | ndb, POWER8, PowerPC |
[13 Nov 2014 3:47]
Stewart Smith
[2 Feb 2015 7:32]
Stewart Smith
This test passes with 7.3.8 and on *LITTLE* endian ppc64. Unless there is a plat to support NDB on any big endian platform (are any current platforms big endian?) it may be better to either close this or just add a configure check to no longer build on big endian platforms.
[5 Mar 2015 11:29]
Mikael Ronström
Pretty sure SPARC is still big endian.
[18 Dec 2015 20:59]
Gustaf Thorslund
Hi Stewart, Using 7.3.7 I've verified this bug almost as described. System used was qemu-system-ppc64 running Debian Jessie. Due to this I also had to update some timeout variables, both for ndbd and mtr. I also had to copy storage/ndb/src/kernel/vm/mt-asm.h from 7.4.9 to get it to compile at all. $ ./mysql-test/mtr ndb.ndb_restore_conv_more |& tee ndb.ndb_restore_conv_more.log . . ndb.ndb_restore_conv_more [ fail ] Test ended at 2015-12-18 16:29:44 CURRENT_TEST: ndb.ndb_restore_conv_more /home/gustaf/src/mysql-cluster-gpl-7.3.7/storage/ndb/tools/ndb_lib_move_data.cpp:1048: require((code != 0)) failed 1043 void 1044 Ndb_move_data::set_error_code(int code, const char* fmt, ...) 1045 { 1046 va_list ap; 1047 va_start(ap, fmt); 1048 require(code != 0); 1049 Error& e = m_error; 1050 e.code = code; 1051 my_vsnprintf(e.message, sizeof(e.message), fmt, ap); 1052 va_end(ap); 1053 } $ tail mysql-test/var/log/ndb.ndb_restore_conv_more/tmp/ndb_testrun.log Processing data in table: test/def/NDB$BLOB_18_3(20) fragment 1 _____________________________________________________ Processing data in table: test/def/NDB$BLOB_18_3(20) fragment 3 _____________________________________________________ Processing data in table: test/def/NDB$BLOB_18_3(20) fragment 5 _____________________________________________________ Processing data in table: test/def/NDB$BLOB_18_3(20) fragment 7 Opening file '/home/gustaf/src/mysql-cluster-gpl-7.3.7/build/mysql-test/var/mysql_cluster.1//BACKUP/BACKUP-8/BACKUP-8.2.log' File size 1024 bytes Restored 590 tuples and 0 log entries $ cat /proc/cpuinfo processor : 0 cpu : POWER7 (raw), altivec supported clock : 1000.000000MHz revision : 2.3 (pvr 003f 0203) timebase : 512000000 platform : pSeries model : IBM pSeries (emulated by qemu) machine : CHRP IBM pSeries (emulated by qemu) If I instead try the current version 7.4.9 I cannot reproduce the bug: ndb.ndb_restore_conv_more [ pass ] 7769970 -------------------------------------------------------------------------- The servers were restarted 0 times Spent 7769.970 of 8369 seconds executing testcases Completed: All 1 tests were successful. real 139m29.313s user 46m59.335s sys 91m31.128s Could you try the current GA too? /Gustaf
[19 Jan 2016 1:00]
Bugs System
No feedback was provided for this bug for over a month, so it is being suspended automatically. If you are able to provide the information that was originally requested, please do so and change the status of the bug back to "Open".