Memory Limits New from S 39.5.11

Objectives

• Limitability of the memory used by the PLANTA Server
• Treatment of memory bottlenecks
• Maintenance of the work ability also for scarce resources

Configuration

• In the delivery condition, the memory restriction is switched off.
• The parameters of the memory restriction are adjustable via the globals.conf configuration file.
• The activation is done by specifying a maximum memory size via the mem_limit.max_size parameter.
  • This size is to be selected depending on the system resources.
  • Here, you have to take into account that the size of the Java Virtual Machine is not included and has to be pulled off from the system resources.
  • Furthermore, further resources which are not included either (approximately 1.5 GiB for 50 parallel sessions), are allocated in each active session.
  • PLANTA recommends for example to use mem_limit.max_size=8192 for a server with 16 GiB RAM , and “medium” for system dimensioning (hence a JMV configuration of 4096 MiB).
• The specifications for required free memory for particular actions should only be adjusted after consulting PLANTA, since a change may have negative effects on the operation of the PLANTA server.

Effects of Memory Shortage

Information

• In the case of memory shortage, the following messages are possibly displayed:

Dialog message 1045 (The action "[action name]" is canceled due to memory shortage.)

Dialog message 1046 (Your session is canceled due to acute memory shortage.)

• Here, the message will look similar to the following one:

• Due to an error, the message now possibly does not contain any text and is therefore displayed in the following way:

Python Exception(s) Memory limit exceeded!

• If memory shortage occurs in the Python context (e.g. from a macro), the following message is output if the exception remains untreated:

Situations and Effects

Information

• In the actions mentioned below, it is checked whether the estimated memory size is still available.
• If this is not the case, the action is canceled.
• The user is informed about the cancellation and the reason for it.
• Each action cancellation has specific effects, which are explained in the following.

Start New Session

• The check will be carried out before the user logs off.
• Message 1045 is displayed.
• After confirming via OK, the session is closed.

Load Modules

• The check is done each time the module data is loaded (e.g. filter).
• Message 1045 is displayed.
• If the loading of the module data is initiated from the Python code, a Python exception is also output.
• The loading of the module data (analogously to the cancellation of the module filter via ESC) is canceled.

Calculate Project

• The check is done before the calculation of a project.
• Message 1045 is displayed.
• A Python exception is output if the Python code calculation has been called.
• The calculation is not carried out.

Replanning

• The check is done before replanning is started.
• Message 1045 is displayed to the user.
• A Python exception is output if the Python code calculation has been called.
• The calculation is not carried out.

Overrun of the Allocation Limit

• The check is done for allocations.
• Message 1046 is displayed.
• Depending on the action performed, there may also be Python exceptions.
• The session is quit after the message is confirmed.