[rt-users] RT using 100% cpu when opening some tickets

[Byron Jones]

when we open some tickets on RT 3.6.0 (plan to upgrade to 3.6.1 when i
can) the httpd process uses 100% CPU and never stops.
 
here's what top looks like with two requests to open the same ticket:
 
  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
17556 httpdnin  25   0  423m 406m 2760 R 99.9 26.8   8:20.84 httpd
20343 httpdnin  25   0  391m 372m 2624 R 99.6 24.6   2:51.43 httpd
20429 root      16   0  1968 1108  828 R  0.3  0.1   0:00.04 top
...
 
here's the mysql process list:
 
|   3 | rt_nintek | localhost | rt_nintek | Query   |  290 | User lock |
SELECT GET_LOCK('Apache-Session-8dc6987e94c140bf84fb5fe379594613', 3600)
|
|   5 | rt_nintek | localhost | rt_nintek | Query   |  280 | User lock |
SELECT GET_LOCK('Apache-Session-8dc6987e94c140bf84fb5fe379594613', 3600)
|
|   6 | rt_nintek | localhost | rt_nintek | Query   |  290 | User lock |
SELECT GET_LOCK('Apache-Session-8dc6987e94c140bf84fb5fe379594613', 3600)
|
|   7 | rt_nintek | localhost | rt_nintek | Query   |  203 | User lock |
SELECT GET_LOCK('Apache-Session-db0718c787b8a257d22087a9a99fb51a', 3600)
|
|   8 | rt_nintek | localhost | rt_nintek | Sleep   |  193 |           |
NULL
|
|   9 | rt_nintek | localhost | rt_nintek | Query   |  379 | User lock |
SELECT GET_LOCK('Apache-Session-8dc6987e94c140bf84fb5fe379594613', 3600)
|
|  10 | rt_nintek | localhost | rt_nintek | Query   |  413 | User lock |
SELECT GET_LOCK('Apache-Session-8dc6987e94c140bf84fb5fe379594613', 3600)
|
|  11 | rt_nintek | localhost | rt_nintek | Sleep   |  193 |           |
NULL
|
|  13 | rt_nintek | localhost | rt_nintek | Query   |  203 | User lock |
SELECT GET_LOCK('Apache-Session-db0718c787b8a257d22087a9a99fb51a', 3600)
|
| 399 | rt_nintek | localhost | rt_nintek | Query   |  236 | User lock |
SELECT GET_LOCK('Apache-Session-8dc6987e94c140bf84fb5fe379594613', 3600)
|
| 400 | rt_nintek | localhost | rt_nintek | Query   |  234 | User lock |
SELECT GET_LOCK('Apache-Session-8dc6987e94c140bf84fb5fe379594613', 3600)
|
| 401 | rt_nintek | localhost | rt_nintek | Query   |  230 | User lock |
SELECT GET_LOCK('Apache-Session-8dc6987e94c140bf84fb5fe379594613', 3600)
|
| 402 | rt_nintek | localhost | rt_nintek | Query   |  203 | User lock |
SELECT GET_LOCK('Apache-Session-db0718c787b8a257d22087a9a99fb51a', 3600)
|
| 403 | rt_nintek | localhost | rt_nintek | Sleep   |    2 |           |
NULL
|
| 411 | rt_nintek | localhost | rt_nintek | Sleep   |  127 |           |
NULL
|
| 413 | rt_nintek | localhost | rt_nintek | Sleep   |  127 |           |
NULL
|
| 415 | rt_nintek | localhost | rt_nintek | Sleep   |   47 |           |
NULL
|
| 416 | rt_nintek | localhost | rt_nintek | Sleep   |    0 |           |
NULL
|
| 417 | rt_nintek | localhost | rt_nintek | Sleep   |    1 |           |
NULL
|
| 418 | rt_nintek | localhost | rt_nintek | Sleep   |    1 |           |
NULL
|

help!
 
it's at the point now where i have a script that runs every minute
killing httpd processes that have been running for a long time with high
cpu usage.

 

[ byron jones
[ zoom itg                 "In 3010, the potatoes triumphed"
[ phone +61 8 9389 5235                        - N. Gertler



 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.bestpractical.com/pipermail/rt-users/attachments/20060906/626eee57/attachment.htm>