perfSONAR User Q&A and Other Discussion

[Mark Feit <>]

Whitworth, Luke writes:

 

One final bit on this, everything seems quite happy with ASD running, only thing is the tables run and archiving in the pscheduler DB seem to grow and grow:

 

    table_schema    |       table_name        | row_estimate |   total    |   index    |   toast    |   table   

--------------------+-------------------------+--------------+------------+------------+------------+------------

public             | run                     |        13214 | 59 MB      | 40 MB      | 5576 kB    | 13 MB

public             | archiving               |        13140 | 17 MB      | 7784 kB    | 8192 bytes | 9296 kB

 

Is there a safe way to truncate these periodically……or is there a flag to bypass them completely?

 

Both tables are where 99% of pScheduler’s work is done and are, therefore, important to its functioning.

 

Assuming a relatively-steady workload, the tables on a fresh system will grow until reaching a steady state.  pScheduler drops runs and the archivings tied to them) 48 hours after completion, so how big the table gets depends on how many runs happen during that period.

 

The amount of time runs are kept is an internally-tunable knob that hasn’t been brought to the outside.  If you want to experiment with it, drop me a line off-list and I’ll describe how to change it on a if-you-break-it-you-own-both-pieces basis.

 

--Mark