Bug #47965 Cluster Replication - segfault on startup
Submitted: 10 Oct 2009 0:34 Modified: 12 Nov 2009 13:09
Reporter: John David Duncan Email Updates:
Status: No Feedback Impact on me:
None 
Category:MySQL Cluster: Replication Severity:S3 (Non-critical)
Version:mysql-5.1-telco-7.0 OS:Any
Assigned to: CPU Architecture:Any
Tags: 7.0.8a

[10 Oct 2009 0:34] John David Duncan
Description:
After restoring some data from an ndb backup, it was noticed that two non-replicating sql nodes ran fine but two replication nodes crashed on startup with a segfault.  It could be that the stored ndb_binlog information was inconsistent with the state of the database after the restore.  

How to repeat:
intermittent issue. unable to reproduce consistently.
[10 Oct 2009 0:43] John David Duncan
Segfault in ndbcluster_binlog_index_purge_file  [ line 568]:

  ---- called from signal handler with signal ??? ------
  [7] ndbcluster_binlog_index_purge_file(thd = ???, file = ???) (optimized), at 0x1004252d8 (line ~568) in "ha_ndbcluster_binlog.cc"
  [8] ndbcluster_binlog_func(hton = ???, thd = ???, fn = ???, arg = ???) (optimized), at 0x10042620c (line ~961) in "ha_ndbcluster_binlog.cc"
  [9] binlog_func_foreach(thd = ???, bfn = ???) (optimized), at 0x1003766e4 
  [10] ha_binlog_index_purge_file(thd = ???, file = ???) (optimized), at 0x100376860 
  [11] MYSQL_BIN_LOG::purge_logs(this = ???, to_log = ???, included = ???, need_mutex = ???, need_update_threads = ???, decrease_log_space = ???) (optimized), at 0x10032dd18 
  [12] MYSQL_BIN_LOG::purge_logs_before_date(this = ???, purge_time = ???) (optimized), at 0x10032e480 
  [13] init_server_components() (optimized), at 0x100280f44 
  [14] main(argc = ???, argv = ???) (optimized), at 0x100281388
[12 Oct 2009 8:34] Jonas Oreland
JD,

Was the replicating mysqld running while ndb_restore was running?
Did the ndb_restore run restore blobs ?

In this case this is a known "not supported"
Not 100% where to find this in documentation...
[13 Nov 2009 0:00] 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".
[9 Nov 2012 7:52] MySQL Verification Team
JDD did you have expire_logs_days ??