Generic x86_64 (GPT) performance issue

Hello,
During my recent evaluation of balenaOS on the IPC, I encountered an issue that has raised concerns regarding its performance. Upon deploying the OS and initiating operations, I observed a significant degradation in system responsiveness and speed. This issue was particularly evident in the high virtual memory usage, as indicated by the output of the “top” command provided below:

Mem: 884052K used, 7092276K free, 10552K shrd, 53092K buff, 489932K cached
CPU: 0% usr 0% sys 0% nic 100% idle 0% io 0% irq 0% sirq
Load average: 0.00 0.00 0.00 1/186 8746
PID PPID USER STAT VSZ %VSZ %CPU COMMAND
8746 8741 root R 4144 0% 0% top
1755 1 root S 1531m 19% 0% /usr/bin/balenad --experimental --log-driver=journald --storage-driver=overlay2 -H fd://
2815 1 root S 1530m 19% 0% /usr/bin/balena-containerd-shim-runc-v2 -namespace moby -id ee566c4114afd11dd6d8f4565c85
1821 1755 root S 1459m 19% 0% balena-engine-containerd --config /var/run/balena-engine/containerd/containerd.toml --lo
2798 2730 root S 1314m 17% 0% balena start --attach balena_supervisor
2835 2815 root S 324m 4% 0% node /usr/src/app/dist/app.js
1656 1 root S 314m 4% 0% /usr/sbin/NetworkManager --no-daemon
1250 1 root S 307m 4% 0% /usr/sbin/ModemManager --log-journal
1 0 root S 156m 2% 0% {systemd} /sbin/init
2332 1 root S 77820 1% 0% /usr/sbin/chronyd -d

Could you please help me figure this out? Is this performance level correct? If not, I would like to understand the root cause of the issue.

Hello @Vladimir first of all welcome to the balena community!

Could you please confirm what balenaOS and supervisor versions are you using?

Hello @mpous! Thank you for the warm welcome to the balena community. I am using the latest versions: balenaOS 5.1.50+rev1 and supervisor 16.1.0.

What device type are you using @Vladimir ?

I am using Omron NY-series IPC. Generic x86_64 (GPT) HW type @mpous.

@Vladimir And this is happening only to 1 device or to more than 1 device on your fleet?

@mpous I’ve just started to use balena, so I have only 1 device in my fleet right now. I’m using this IPC for ST balena’s device management solution.

Dear @mpous,
At Omron we are evaluating what a partner can provide in the area of IoT connections and platforms. Therefore we downloaded Generic x86_64 (GPT) image. During our evaluation, we have observed slow response of the system and identified high memory consumption. Is it possible to investigate with somebody from your team, what’s happening? You can contact me via email vladimir.iurchenko@omron.com.
Best regards,
Vladimir Iurchenko
SW Engineer Omron Europe

1 Like

@Vladimir my colleagues will contact you via e-mail!