Zabbix server query failed: Lock wait timeout exceeded.

[mtggit]

I installed the new Zabbix server 1.4.1 and port all the hosts to the new server. Initially the server was working perfectly. However, just one week later, the server, the server failed to capture data very frequently. I got the error message as follows:

12985:20070730:164715 Query failed: [insert into history (clock,itemid,value) v
alues (1185785183,18938,0.518657)] Lock wait timeout exceeded; try restarting tr
ansaction [1205]
12982:20070730:164757 Query failed: [insert into history_uint (clock,itemid,val
ue) values (1185785226,19080,892775915)] Lock wait timeout exceeded; try restart
ing transaction [1205]
12986:20070730:164758 Query failed: [insert into trends (clock,itemid,num,value
_min,value_avg,value_max) values (1185782400,20219,1,0.000000,0.000000,0.000000)
] Lock wait timeout exceeded; try restarting transaction [1205]

I read some posts in the forum and found the problem is related to the hourly housekeeping tasks. In fact, seems no data can be captured during the housekeeping period. Please kindly help.

Configuration:

SuSE SLES10
mysql server 5.0.18 default installation with default innodb options
Total hosts = 24

[aNt1_v!RuS]

At me the same problem.
Zabbix v1.4.1
~50 servers
MySQL
24394:20070730:150607 Query failed: [insert into history (clock,itemid,value) values (1185793516,100000000020226,100.000000)
] Lock wait timeout exceeded; try restarting transaction [1205]
24393:20070730:150608 Query failed: [insert into history_uint (clock,itemid,value) values (1185793517,100000000020105,120600
055808)] Lock wait timeout exceeded; try restarting transaction [1205]
24395:20070730:150608 Query failed: [insert into history_uint (clock,itemid,value) values (1185793517,100000000020287,425245
9008)] Lock wait timeout exceeded; try restarting transaction [1205]
24400:20070730:150608 ZBX_GET_ACTIVE_CHECKS: host is null. Ignoring.
24399:20070730:150608 ZBX_GET_ACTIVE_CHECKS: host is null. Ignoring.
24398:20070730:150609 ZBX_GET_ACTIVE_CHECKS: host is null. Ignoring.
24397:20070730:150610 Query failed: [insert into history_uint (clock,itemid,value) values (1185793519,100000000018459,6)] Lo
ck wait timeout exceeded; try restarting transaction [1205]
24402:20070730:150610 ZBX_GET_ACTIVE_CHECKS: host is null. Ignoring.
24396:20070730:150611 Query failed: [insert into history (clock,itemid,value) values (1185793520,100000000020168,67.575683)]
Lock wait timeout exceeded; try restarting transaction [1205]

[aNt1_v!RuS]

UPD:
| 65573 | zabbix | 10.1.0.160:55110 | db_zabbix | Query | 0 | updating | delete from history_uint where itemid=100000000019339 limit 500 |

Update Occurs long....

[mtggit]

I found that the database is already 3GB for just 2 weeks. Seems that's why it is so slow. Is it really normal to have so much data? I just use standard Linux template and create my own SNMP monitoring templates.