[BUG] issue with last arm docker images

[vdiravka]

Describe the bug
All amd versions works fine on my env. For ARM node only v1.21.0-arm64 and v1.21.0-arm64v8 works fine. But v1.26.7-arm64v8 and v1.26.8-arm64v8 (or multi-platform v1.26.8) versions don't running within the same node:

kubectl logs -f --tail=200 dsp-chproxy-0
exec /chproxy: exec format error

To Reproduce
regular chproxy kubernetes deployment by using images from https://hub.docker.com/r/contentsquareplatform/chproxy/tags

Expected behavior
chproxy pod is running (like for v1.21.0-arm64 version)

Screenshots
no need

Environment information
Node: Taints: kubernetes.io/arch=arm64:NoSchedule, Clickhouse 23.10.4.25 version

Additional context
The assumption is #503 not fully tested.
Any suggestions about possible fix? If needed, I will consider the contribution.

[mga-chka]

thanks for the warning, we'll take a look

[jb-abbadie]

Sorry for the disturbance. The PR was checked on a Mac M1 which hid the issue because of transparent Rosetta translation layer...
I managed to reproduce the bug and fix it with #520

You should be good using version starting v1.26.9

[vdiravka]

Thank you @jb-abbadie @mga-chka for quick solution. The new v1.26.9 image works fine.