Basic Benchmarking

Registered by Kurt Griffiths

This is the first iteration on our automated performance testing blueprint.

We need to set up a simple, "good enough for now" benchmarking cluster. Also, we need to create a simple "marconi-bench" tool that will have two clients, one producer and one consumer.'' There will be a "conductor" to start the two clients and collect the results. We will write results out for gnuplot, and also monitor real-time stats via graphite to catch anomalies.

Cluster:

1. HAProxy load balancer
2. 4x web heads running uwsgi, with Keystone auth disabled (we will bench it enabled later)
3. 1x3 mongo replica set
4. 1x graphite box
5. 3x Load generator boxes (1x producer, 1x consumer, one conductor)

Blueprint information

Status:
Complete
Approver:
Malini Kamalambal
Priority:
Medium
Drafter:
Kurt Griffiths
Direction:
Approved
Assignee:
Sriram
Definition:
Approved
Series goal:
Accepted for juno
Implementation:
Implemented
Milestone target:
milestone icon 2014.2
Started by
Kurt Griffiths
Completed by
Kurt Griffiths

Related branches

Sprints

Whiteboard

Gerrit topic: https://review.openstack.org/#q,topic:bp/basic-benchmarking,n,z

Addressed by: https://review.openstack.org/98875
    feat(benchmarking) : Producer-Consumer scenario

(?)

Work Items

Dependency tree

* Blueprints in grey have been implemented.

This blueprint contains Public information 
Everyone can see this information.

Subscribers

No subscribers.