MySQL Bugs: #12222: One instance of TPC-B is getting Lock wait timeout exceeded (new cluster drop)

Bug #12222	One instance of TPC-B is getting Lock wait timeout exceeded (new cluster drop)
Submitted:	27 Jul 2005 18:17	Modified:	23 Sep 2005 16:13
Reporter:	Jonathan Miller	Email Updates:
Status:	Closed	Impact on me:	None
Category:	MySQL Cluster: Cluster (NDB) storage engine	Severity:	S2 (Serious)
Version:	5.1-wl2325	OS:	Linux (Linux)
Assigned to:	Pekka Nousiainen	CPU Architecture:	Any

Description:
I had only been getting this under heavy load on the clone before this clone. Now I am getting this for just one instance of the test.

[ndbdev@ndb09 scripts]$ . ./run_tpcb.sh
Total transaction for ndb09_1 completed so far = 1000
Total transaction for ndb09_1 completed so far = 2000
Total transaction for ndb09_1 completed so far = 3000
Total transaction for ndb09_1 completed so far = 4000
DBD::mysql::st execute failed: Lock wait timeout exceeded; try restarting transaction at ./TPCB.pl line 278.
Update Account execute failed
1 at ./TPCB.pl line 278.
Transaction failed and was rolled back!!!
Total transaction for ndb09_1 completed so far = 5000
DBD::mysql::st execute failed: Lock wait timeout exceeded; try restarting transaction at ./TPCB.pl line 288.
Update Teller execute failed
1 at ./TPCB.pl line 288.
Transaction failed and was rolled back!!!
Total transaction for ndb09_1 completed so far = 6000
Total transaction for ndb09_1 completed so far = 7000
Total transaction for ndb09_1 completed so far = 8000
Total transaction for ndb09_1 completed so far = 9000
DBD::mysql::db commit failed: Lock wait timeout exceeded; try restarting transaction at ./TPCB.pl line 306.
Total transaction for ndb09_1 completed so far = 10000
Total transaction for ndb09_1 completed so far = 11000
DBD::mysql::st execute failed: Lock wait timeout exceeded; try restarting transaction at ./TPCB.pl line 278.
Update Account execute failed
1 at ./TPCB.pl line 278.
Transaction failed and was rolled back!!!
Total transaction for ndb09_1 completed so far = 12000
Total transaction for ndb09_1 completed so far = 13000
Total transaction for ndb09_1 completed so far = 14000
Total transaction for ndb09_1 completed so far = 15000
DBD::mysql::st execute failed: Lock wait timeout exceeded; try restarting transaction at ./TPCB.pl line 238.
Select Branch execute failed
1 at ./TPCB.pl line 238.
Transaction failed and was rolled back!!!
Total transaction for ndb09_1 completed so far = 16000
DBD::mysql::st execute failed: Lock wait timeout exceeded; try restarting transaction at ./TPCB.pl line 249.
Select teller execute failed
1 at ./TPCB.pl line 249.
Transaction failed and was rolled back!!!
Total transaction for ndb09_1 completed so far = 17000
Total transaction for ndb09_1 completed so far = 18000
Total transaction for ndb09_1 completed so far = 19000
Total transaction for ndb09_1 completed so far = 20000
Total transaction for ndb09_1 completed so far = 21000
DBD::mysql::st execute failed: Lock wait timeout exceeded; try restarting transaction at ./TPCB.pl line 238.
Select Branch execute failed
1 at ./TPCB.pl line 238.
Transaction failed and was rolled back!!!

How to repeat:
run TPC-B against 3 host cluster with replication.

do you get this only if replication is on, or is enough with just having the binlog updated.

If so, is this reproducable even without the binlog turned on?

Also, can you reproduce in a smaller setup, e.g. running the whole app, including cluster, on a single host?

do you get this only if replication is on, or is enough with just having the binlog updated.

* Not sure

If so, is this reproducable even without the binlog turned on?

* Not Sure

Also, can you reproduce in a smaller setup, e.g. running the whole app, including cluster, on a single host?

*Not sure, but it would not take much to try it.

This was a change in results going from one clone to another. One instance of the application should not be getting this. I am not totally sure what changed.

Found and fixed bugs with test scripts