Case study - load on a server launching 100 $Maria masternodes (in 2 batches of 50)

[HelmutsHelmuts is verified member.]

This is a case study what load is there on a server when you plan to host 100 $Maria masternodes on it.

I am planning to launch 50 masternodes at once, monitor the process, and after that launch the next 50 masternodes.

Quick note about the server:
CPU: Intel Xeon E3-1230v6 - 4c/8t - 3.5 GHz/3.9 GHz
RAM: 64 GB ECC 2133 MHz

--

Ok, let's go.

3 screenshots for a person who is wondering what is a load on a server launching 50 $Maria masternodes.

Just restarted the server, launched first 50 masternodes and now I am monitoring the server loads using 'htop'.

The idea is to launch the next 50 only only after the 1st batch of address re-syncing is over (will be judging from htop).

 

[HelmutsHelmuts is verified member.]

Ok, nice, it seems the processor cores have done their job, now the rest is on the ram:

 

[HelmutsHelmuts is verified member.]

ok, started next 50 masternodes..

start looks clean.. need to wait more

 

[HelmutsHelmuts is verified member.]

ok.. now the loads start appearing (up and down):




Need to wait till the process is over.

Interesting note: Swap already has been used. :/ .. bad sign

 

[HelmutsHelmuts is verified member.]

The process of the 2nd batch seems to be over:



We can see that RAM is 90% used.

So, without optimising the masternodes for consuming less RAM > this should be where you draw a line (don't add more masternodes).

Current settings from the maria.conf (part of the file, of course):

maxconnections=256
addnode=188.40.233.38
addnode=188.40.233.39
addnode=188.40.233.40
addnode=188.40.233.41
addnode=188.40.233.43
addnode=188.40.233.44
addnode=188.40.233.45
addnode=199.127.140.224
addnode=199.127.140.225
addnode=199.127.140.235
addnode=199.127.140.236
addnode=94.130.95.106

.. this case study will be continued as I will be exploring and testing approaches saving RAM while running masternodes.  ... will be shared here.

Have a great day! Helmuts