You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Document hardware requirements for running the samples locally #36
The README doesn't mention hardware requirements yet. Some numbers from running scripts/01-system-assigned-managed-identity.sh (Windows 11, WSL, Ubuntu) with 1 system and 1 user node.
docker stats:
Container
Memory
k3d server (control plane)
1.2 GB
system node
0.9 GB
localstack-azure
0.6 GB
user node
0.5 GB
azurite + k3d load balancer
0.08 GB
So ~3.3 GB for the containers, roughly 0.5–1 GB per additional node. docker system df: ~6 GB images, ~7.6 GB volumes.
On WSL, vmmemwsl grew to ~18 GB. That turned out to be mostly page cache from unpacking the images (free -h: 4.4 GB used, 13 GB buff/cache), which WSL doesn't return to Windows quickly, so it looks much heavier than it is.
Suggestions:
Add minimum RAM and disk space to the README prerequisites.
Mention the approximate memory per node in the node-count prompt.
For WSL users, a tip on autoMemoryReclaim or a memory= limit in .wslconfig.
We had the same idea: we planned to run some tests and document the minimum configuration in terms of RAM and disk space to run the samples. So far, we didn't find the time to do it, but this is definitely something in our todo list. I'll close this issue for now, and create an issue in our project management and issue tracking system!
Activity
pdebruin commentedon Oct 14, 2026
Tagging @paolosalvatori FYI
paolosalvatori commentedon Oct 14, 2026
We had the same idea: we planned to run some tests and document the minimum configuration in terms of RAM and disk space to run the samples. So far, we didn't find the time to do it, but this is definitely something in our todo list. I'll close this issue for now, and create an issue in our project management and issue tracking system!
paolosalvatori commentedon Oct 14, 2026
@pdebruin I used your data to create #43 PR. We'll continue to evaluate system requirements for other tests. Thanks for the feedback! cc: @lazarkanelov, @mmaureenliu, @DrisDary, @bryansan-local