-
Notifications
You must be signed in to change notification settings - Fork 74
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
Baremetal e2e scripts #248
Conversation
* Updated workloads to run on baremetal * first commit: skipping cleanup on baremetal * first commit: skipping cleanup on baremetal * fixed spacing * Checking if cloud is on bareMetal. If it is, we're no longer deleting the benchmark-operator namespace in between runs Co-authored-by: jdowni000 <[email protected]>
* first commit: adding calculations for allocatable CPU per MCP * Changing approach to creating new n number mcps depending on node count * removing old machineConfig_pool func * removing bash script testing code left in common.sh * fixed a couple naming convention issues * small fixes from testing in actual cluster * fixed applying mcp.yaml to use envsubst * more issues resolved from testing * more fixes in while loops while testing * completed deployment fix * changing function name, adding check for mcps and ns if exists * adding json creation and use of mb * adding logic to allow MCP_SIZE and MCP_NODE_COUNT to be set as a variable * changing env var MCP_SIZE to TOTAL_MCPS * fixed spacing in all of new function * adding mb_pod.yml and sending mb operations to pod instead of cli * removing response.csv file as it may be too large depending on how many mcps are genereated * added var to set to choose whether MCPs are created or not * just fix...its Friday...its fixed * fixed if statemnet with elif * creating logic to check env var inputs * adding if condition for sleep depending on size of node_count * adding resources to mb-pod * adding logic to check for sample app pods to be ready before mb-pod deployment * Update common.sh mb-pod and sample app
rerun all |
/rerun all |
Results for e2e-benchmarking CI Tests
|
workloads/network-perf/common.sh
Outdated
|
||
#If using baremetal we use different query to find worker nodes | ||
if [[ "${isBareMetal}" == "true" ]]; then | ||
log "Colocating uperf pods for baremetal" |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We are not really colocating are we?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
right, we are not. we randomly pick 2 worker nodes. I will delete this log.
workloads/kube-burner/common.sh
Outdated
kubectl apply -f benchmark-operator/resources/kube-burner-role.yml | ||
log "Waiting for benchmark-operator to be running" | ||
oc wait --for=condition=available "deployment/benchmark-controller-manager" -n benchmark-operator --timeout=300s | ||
if [[ "${isBareMetal}" == "false" ]]; then |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
tbh i think we should just leave the operator around each time and change this to apply to all cluster types. I don't see a reason why we need to delete/recreate the operator
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
+1
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Once the operator is deployed, we shouldn't delete it.
I'd modify the current code to always try to deploy the operator. With the current make deploy
implementation, if the operator is already running it won't be redeployed.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Removed delete command
2580f8b
to
137526c
Compare
137526c
to
fdb68a6
Compare
@mohit-sheth @rsevilla87 @whitleykeith any more suggestions/comments ? |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
lgtm
/rerun all |
Results for e2e-benchmarking CI Tests
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
@rsevilla87 fixed conflicts and tested, ptal. |
Signed-off-by: Raul Sevilla <[email protected]>
Signed-off-by: Raul Sevilla <[email protected]>
/rerun all |
Results for e2e-benchmarking CI Tests
|
/rerun all router-perf-v2:ingress-performance.sh |
Results for e2e-benchmarking CI Tests
|
Changes for baremetal Co-authored-by: jdowni000 <[email protected]> Co-authored-by: Marko Karg <[email protected]> Co-authored-by: jdowni000 <[email protected]> Co-authored-by: Raul Sevilla <[email protected]>
Description
Fixes
#208
Developed by @jdowni000, I just fixed few things and corrected conflicts as he is on PTO. PR #225 can be marked as obsolete.