Bug #57558 4 ndbmtds sometimes never enter startphases above 4

• Submitted: 19 Oct 2010 9:38
• Modified: 7 Mar 2011 12:24
• Reporter: Thomas Nielsen
• Status: No Feedback
• Category: MySQL Cluster: Cluster (NDB) storage engine
• Severity: S2 (Serious)
• Version: mysql-5.1-telco-7.1
• OS: Linux (rhel5)
• CPU Architecture: Any
• Tags: 7.1.8

[19 Oct 2010 9:38] Thomas Nielsen

Description:
Cluster with 4 ndbmtds sometimes never get to startphase >4 during initial start

How to repeat:
create cluster
  #1 mgmd
  #2 ndbmtd
  #3 ndbmtd
  #4 ndbmtd
  #5 ndbmtd
  #6 mysqld
change config variables according to config.ini/my.cnf
start cluster   <-- startphase 4 completed is last output

After 5 minutes the test decides enough is enough, and kills off the cluster.

[20 Oct 2010 12:33] Jonas Oreland

why do you have those weird [TCP] sections ?
what happens if you remove them.
where are out files from data nodes ?

/Jonas

[7 Feb 2011 12:24] Jonas Oreland

there is really not enough info for this,
can I ask that you reproduce this with
1) latest 7.1 (10)
2) with maximum log level
  (ndb_mgm -e "all clusterlog XXX=15") for each legal value of XXX
  *before* starting ndb(mt)d

[8 Mar 2011 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".