Skip to content

Latest commit

 

History

History
78 lines (57 loc) · 3.56 KB

ycsb.md

File metadata and controls

78 lines (57 loc) · 3.56 KB
Raw

YCSB

YCSB is a performance test kit for key-value and other cloud serving stores.

Running YCSB

Given that you followed instructions to deploy operator, you can modify cr.yaml to your needs.

YCSB is a workload that requires a database/key-value store to run workloads against and benchmark.

Your resource file may look like this:

apiVersion: ripsaw.cloudbulldozer.io/v1alpha1
kind: Benchmark
metadata:
  name: ycsb-mongo-benchmark
  namespace: benchmark-operator
spec:
  clustername: myk8scluster
  workload:
    name: ycsb
    args:
      infra: mongodb
      driver: mongodb
      recordcount: 100
      operationcount: 100
      workloads:
        - workloada
        - workloadb
      options_load: '-p mongodb.url="mongodb://mongo/ycsb?"' #passed as is to ycsb when loading database
      options_run: '-p mongodb.url="mongodb://mongo/ycsb?" -threads 10 -target 100'
      pin: false
      #pin_node:

The following options in args are required:

infra is the database against which you're running mongodb

driver is the driver used to interact with database, you can choose from YCSB Github

recordcount number of records to be inserted into the database (insertcount defaults to recordcount).

operationcount number of operations to run against the database

workload is a list of workloads to run, you can read about the workloads at YCSB workloads Note: We currently support running either workloade or workloadd( and not in the same list) as the last workload in the list. If you'd like to run both of them, then you'll have to apply the cr with workloadd/workloade as the last workload in the args. Then after it finishes running all workloads, you'll have to manually drop the database. Then create another CR with a different benchmark name with just the workloade/workloadd in the workload list that wasn't run previously. Reason: Similar to reason provided by YCSB documentation, we'd need to drop the database after running either of the workloadd/workloade. This would mean operator to have knowledge of interaction with the said DB, which was outside the scope of this operator.

options_run these are the options passed to ycsb binary while running the workload, this is where the target database and options are passed. Please read YCSB documentation on the necessary args required for the particular database. And note, the url or API needs to be accessible from the ycsb pod.

The following options are optional:

loaded This can be optionally set to true, if you have already loaded the database.

options_load This needs to be set if loaded is not defined or set to false, and like in the case of options_run this needs to be configured properly, so that the ycsb pod can access the API of database.

pin will allow the benchmark runner place nodes on specific nodes, using the hostname label.

pin_node what node to pin the pod to.

runtime_class: If this is set, the benchmark-operator will apply the runtime_class to the podSpec runtimeClassName.

annotations If this is set, the benchmark-operator will set the specified annotations on the pods' metadata.

Once done creating/editing the resource file, you can run it by:

# kubectl apply -f config/samples/ycsb/cr.yaml # if edited the original one
# kubectl apply -f <path_to_file> # if created a new cr file