System administration

Question

Hi,

I wondered if you had any recommendation about system (nodes) monitoring. What are the main health parameters we must keep an eye on, except memory, disk usage or any other regular "OS level" monitoring ?

Does Multichain provide a "regular" logfile showing potential errors? (without using "debug")

Thanks and congrats for this exciting technology (y)

Answer 1

First, the debug.log file is still output with a lot of diagnostic information, even if you have not set any -debug flags.

Other things to monitor:

• The utxocount in getwalletinfo to not go above a few 10000s, at which point new transaction creation will slow down significantly.
• The number of transactions and their byte size in getmempoolinfo should not be too high for what you expect - this would signify your node has become disconnected or blocks have stopped being created.
• Various fields of the output of getpeerinfo that indicate how many peers are connected, their long times, etc...

And yes, you should also monitor memory, CPU and disk usage at the OS level.