Skip to content
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

[WIP] Add load balancer support. #218

Open
wants to merge 2 commits into
base: master
Choose a base branch
from

Conversation

dceara
Copy link
Collaborator

@dceara dceara commented Jun 3, 2021

No description provided.

This is still WIP because we have to find a way to not reparse all load
balancers from the NB DB every time.

Signed-off-by: Dumitru Ceara <dceara@redhat.com>
Signed-off-by: Dumitru Ceara <dceara@redhat.com>
@dceara
Copy link
Collaborator Author

dceara commented Jun 3, 2021

@putnopvut It would be great if you could have a look at this, thanks!

@@ -418,6 +418,11 @@ def setup_switch_per_node(self, fake_multinode_args = {},
lport_create_args = {},
port_bind_args = {},
create_mgmt_port = True):
# TODO: figure out how to not reload the context?
Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@putnopvut This is the hack I was talking about.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can you go into a bit more detail about what's going on here? I've been away from this for long enough that I don't even know what it means to reload the context.

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

self.context gets populated once before any iteration of the scenario is run, in class OvnNorthboundContext.setup(). Then, if I'm not wrong, each scenario gets a copy of the context. However, because we're updating load balancers in each iteration, ideally we would have to update the original context, such that upcoming iterations have the updated set of load balancers.

The rally runner doesn't allow that because iterations might be run in parallel, I presume.

To work around it I explicitly updated self.context["ovn-nb-lbs"] with the results of ovn_nbctl.lb_list().

@dceara dceara changed the title Add load balancer support. [WIP] Add load balancer support. Jun 3, 2021
Comment on lines +187 to +195
tokens = line.split('=')
if len(tokens) == 0:
lb_name = None
continue
lb_vip = tokens[0]
lb_backends = tokens[1] if len(tokens) == 2 else ''
lbs[lb_name] = (lb_vip, lb_backends)
lb_name = None

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If the = token is not found, then split() will return a list containing the whole string as the first item. Therefore, len(tokens) wont' ever be 0.

You can just replace this whole block with

lbs[lb_name] = tuple(line.split('=', 2))
lb_name = None

This will ensure that lbs[lb_name] is set to a tuple of 2 items. If the = is not present, then the second item of the tuple will be an empty string.

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Way better, will do, thanks!

self.run("lb-add", [], params)

def lb_set_vip_backends(self, lb_name, lb_vip, lb_backends):
vips_set='vips=\'\"{}\"=\"{}\"\''.format(lb_vip, lb_backends)
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Nit, you can get rid of the \ before the double quotes. It might make the line a bit easier to read.

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Ack, thanks.

@@ -418,6 +418,11 @@ def setup_switch_per_node(self, fake_multinode_args = {},
lport_create_args = {},
port_bind_args = {},
create_mgmt_port = True):
# TODO: figure out how to not reload the context?
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can you go into a bit more detail about what's going on here? I've been away from this for long enough that I don't even know what it means to reload the context.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

2 participants