New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

lxd-benchmark: Change the default count of containers from 100 to 1 #4074

Merged
merged 1 commit into from Dec 5, 2017

Conversation

2 participants
@simos
Contributor

simos commented Dec 5, 2017

The default count for lxd-benchmark is 100 containers.
This can crash systems with less than 16GB of RAM.
The default should be 1, and then let the user select a bigger value.

Signed-off-by: Simos Xenitellis simos@users.noreply.github.com

lxd-benchmark: Change the default count of containers from 100 to 1
The default count for lxd-benchmark is 100 containers.
This can crash systems with less than 16GB of RAM.
The default should be 1, and then let the user select a bigger value.

Signed-off-by: Simos Xenitellis <simos@users.noreply.github.com>
@stgraber

This comment has been minimized.

Show comment
Hide comment
@stgraber

stgraber Dec 5, 2017

Member

Someone reported the same thing in the past. The suggestion was to tweak it to be -1 and then have -1 cause the value to be twice the number of CPU cores or something along those lines.

That should make the benchmark value still vaguely useful out of the box and not run people out of memory or CPU.

Member

stgraber commented Dec 5, 2017

Someone reported the same thing in the past. The suggestion was to tweak it to be -1 and then have -1 cause the value to be twice the number of CPU cores or something along those lines.

That should make the benchmark value still vaguely useful out of the box and not run people out of memory or CPU.

@simos

This comment has been minimized.

Show comment
Hide comment
@simos

simos Dec 5, 2017

Contributor

I think it would be tricky to get it right (to have count > 1) considering all sorts of computers that run Ubuntu. There are octa-core developer boards with microSD storage and there are 96-core (those Caviums) SoCs.
If a user runs lxd.benchmark launch on a production machine with existing containers, they can get in deep trouble.

I propose either,

  • have no default for the number of containers, therefore it is up to the user to make the decision
  • have a default of one container (as other benchmark utilities do).
Contributor

simos commented Dec 5, 2017

I think it would be tricky to get it right (to have count > 1) considering all sorts of computers that run Ubuntu. There are octa-core developer boards with microSD storage and there are 96-core (those Caviums) SoCs.
If a user runs lxd.benchmark launch on a production machine with existing containers, they can get in deep trouble.

I propose either,

  • have no default for the number of containers, therefore it is up to the user to make the decision
  • have a default of one container (as other benchmark utilities do).

@stgraber stgraber merged commit 44a4341 into lxc:master Dec 5, 2017

3 of 4 checks passed

continuous-integration/travis-ci/pr The Travis CI build is in progress
Details
Branch target Branch target is correct
Details
Signed-off-by All commits signed-off
Details
continuous-integration/appveyor/pr AppVeyor build succeeded
Details
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment