node cpu usage

[nikito7]

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
11061 user 20 0 10.760g 85924 62380 S 34.1 8.5 0:28.62 grin

17:44:08 up 10:20, 1 user, load average: 0.47, 0.55, 0.86

Constant high cpu usage, it's normal?

Default configs.

[DavidBurkett]

My mac node sits at ~20% CPU. I assume one of the thread sleep values is unnecessarily low, or one of the containers is really inefficient, but I haven't had a chance to track it down. Grin++ sits at <0.1%, so there's certainly nothing core to the protocol that requires constant high CPU usage.

[nikito7]

changed:

cat .grin/main/grin-server.toml | grep tui
run_tui = false

now <6% cpu usage

[nikito7]

Sooo,

no tui, <8 peers = < 10% cpu
no tui, >50 peers = > 70% cpu

And Grin++, 50 peers = < 5% cpu

[nikito7]

windows preview grin

with tui, 50 peers: 10% cpu
no tui, 50 peers: <1% cpu

[amazarache]

I have followed this thread, trying to achieve less CPU usage for grin node, but without luck :(
disabling tui doesn't lower the CPU usage in my case

processor usage stays between 30-50% load in both cases.

[nikito7]

peer_max_count = 8
peer_min_preferred_count = 4

[amazarache]

Whatever values I used for for peers, grin server loads processor to 40-50%

[nikito7]

Its normal at begin, peers discovery issues

Use default port, 8 peers
And wait a day

[nikito7]

maybe fixed in
20190616 21:08:27.629 INFO grin - This is Grin version 1.1.1-beta.1 (git v1.1.0-3-gd03a81f), built for x86_64-unknown-linux-gnu by rustc 1.35.0 (3c235d560 2019-05-20).

[nikito7]

Current status:
Vps 2Ghz

50% cpu load 275 peers

35% 189 connected (174 most_work). all 5019 = 192 healthy + 6 banned + 4821 defunct / 2.0.0-beta.2

[yeastplume]

Are there any actions to be taken here or can this be closed?

[DavidBurkett]

This was solved long ago. Should be good to close.