Opened 16 years ago
Last modified 16 years ago
#325 new defect
Memory not being freed in fitting
| Reported by: | rwilson | Owned by: | nariman |
|---|---|---|---|
| Priority: | low | Milestone: | Validation of GA tsunami modelling methodology |
| Component: | Efficiency and optimisation | Version: | |
| Severity: | normal | Keywords: | |
| Cc: |
Description (last modified by )
When a simulation run does not find data it can use in the cache it creates the cache data. This process allocates and doesn't free memory, impacting other simulations on the same node. We should investigate the fitting code and explicitly 'unreference' data structures that are maintained in the mesh structure.
This could be done either in existing code that executes when we are finished the fitting, or we could add another method that explicitly frees structures no longer required.
Change History (2)
comment:1 Changed 16 years ago by
| Description: | modified (diff) |
|---|
comment:2 Changed 16 years ago by
| Owner: | changed from ole to nariman |
|---|
Note: See
TracTickets for help on using
tickets.
