
|
If you were logged in you would be able to see more operations.
|
|
|
WET
Created: 19/Oct/06 02:57 AM
Updated: 18/Jun/07 01:56 AM
|
|
| Component/s: |
WET DOM Interface
|
| Affects Version/s: |
0.9.8_beta1,
0.9.8_Patch
|
| Fix Version/s: |
1.2
|
|
|
Original Estimate:
|
Unknown
|
Remaining Estimate:
|
Unknown
|
Time Spent:
|
Unknown
|
|
|
I have noticed that with a constant use of the WET UI, the ruby memory layer seems to have a memory leak - The Ruby process that started off the rpc server initially is at 13 mb or so and after a bit of usage I have seen it go upto 30+ mb. Need to investigate if there is a real memory leak or it is just that a call to a particular rpc method chews up some memory but doesn't do this for every subsequent cal to the same method. We probably some testing to verify this.
In any case it is a good idea to add the following functionality to the RPC Tray Application:
1) Continusoulsy monitor if the rpc server is up and running - this can probably be done once in a minute or so. This can be done using the current ping to rpc server method.
2) Periodically, say once in 10 minutes check for the performance of the rpc tray application - check the memory /cpu usage. Compare it with a predetermined 'max' threshold.
If any of the above fails, then automatically restart the RPC server after showing an error message to the user that reads (something like):
There appears to be a problem with the WET Rpc Server. The RPC Server is being restarted automatically. This should take around a minute. Some parts of WET may not work during this time. If this message doesn't dissappear after 2-3 minutes, please close your system and restart.
|
|
Description
|
I have noticed that with a constant use of the WET UI, the ruby memory layer seems to have a memory leak - The Ruby process that started off the rpc server initially is at 13 mb or so and after a bit of usage I have seen it go upto 30+ mb. Need to investigate if there is a real memory leak or it is just that a call to a particular rpc method chews up some memory but doesn't do this for every subsequent cal to the same method. We probably some testing to verify this.
In any case it is a good idea to add the following functionality to the RPC Tray Application:
1) Continusoulsy monitor if the rpc server is up and running - this can probably be done once in a minute or so. This can be done using the current ping to rpc server method.
2) Periodically, say once in 10 minutes check for the performance of the rpc tray application - check the memory /cpu usage. Compare it with a predetermined 'max' threshold.
If any of the above fails, then automatically restart the RPC server after showing an error message to the user that reads (something like):
There appears to be a problem with the WET Rpc Server. The RPC Server is being restarted automatically. This should take around a minute. Some parts of WET may not work during this time. If this message doesn't dissappear after 2-3 minutes, please close your system and restart.
|
Show » |
|