Re: Decreasing Checkpoints & allowing for heavy load

[Adam D]

Copy and paste that to the docs team! Thats the kind of stuff which is really good in regards to fine tuning / configuring.

Based on those calculations, and having DataMemory of 5000, and IndexMemory of 300 I should be setting the following;
TUP to 200 (207)
and ACC to 12

However I have plain hardware (just SATA disk) so Ive decided to set it to more leniant value of half of what the calculation figures;


# 16 * 4 * 16mb files = 1GB
NoOfFragmentLogFiles = 16

# (4*(2^22)) = 16MB
TimeBetweenLocalCheckpoints = 22

#100 = 100 * 10 * 8kb = 8MBit/sec
#40 = 40 * 10 *8kb = 3.2MBit/sec
NoOfDiskPagesToDiskAfterRestartTUP = 100
NoOfDiskPagesToDiskAfterRestartACC = 40
NoOfDiskPagesToDiskDuringRestartTUP = 160
NoOfDiskPagesToDiskDuringRestartACC = 40

I restarted the cluster after making the config changes and restarting the mgmd, however I didnt really notice much increase.