OpenStack’s big tent has expanded to house new projects: Monasca, Senlin, Fuel, Freezer
and Winstackers. In this series of posts, we’ll give you a quick overview of these projects, introduce you to the project team leads (PTLs) and find out what they’re looking for from the community.

Monasca

What: Monasca is a multi-tenant, highly scalable, performant, fault-tolerant monitoring-as-a-service solution for metrics, complex event processing and logging.

Who: Roland Hochmuth, PTL, is a software architect for Hewlett-Packard Cloud.

What new features/enhancements/fixes are expected for the Mitaka release?
There are a number of new features that are focused on improving the performance of user-interfaces by reducing the amount of client-side processing that needs to be done.

We are in the process of adding a new resource for alarms that will allow the API to query the API and get the total number of alarms grouped and filtered by various criteria, such as the state and severity and dimensions, such as service. See: https://review.openstack.org
/#/c/257607/.
Additionally, the ability to order the list of alarms returned, such as ordering by service, state and severity is being added at: https://review.openstack.org/#/c/260697/
In the Monasca agent, recent enhancements were added to report the status of a KVM VM at: https://review.openstack.org/#/c/240606/ as well as perform a ping test on VMs at: https://review.openstack.org/#/c/258714/.
Support for Monasca in Grafana 2.0 is under active development.
The Monasca Log API is under development at: https://github.com/openstack/monasca-log-api and we expect to have the first official release of the log api available.
An integration of Monasca with Congress is in progress. See: https://review.openstack.org/#/c/241826/.

What contributions do you need most from the community?
One of organizations that is deploying Monasca is really starting to push the boundaries of monitoring-as-a-serice. There are several improvements to the API that we believe could significantly improve performance. One area that seems really compelling though is to develop an in-memory time-series cache for the hot metrics, similar to what Facebook has implemented: http://www.vldb.org/pvldb/vol8/p1816-teller.pdf.
If developers are interested in working in this area, it would be a pretty exciting way to get in on the ground floor.

Get involved!

Use Ask OpenStack for general questions.
For roadmap or development issues, subscribe to the OpenStack development mailing list, and use the tag [monasca]
To get code, report bugs, view blueprints, etc. check out the Monasca Launchpad page
The Monasca team holds weekly meetings Wednesdays at 15:00 UTC in #openstack-meeting-3 on IRC.

Senlin

What: Senlin is a clustering service designed for managing homogeneous objects exposed by any other OpenStack services. It features an open framework where developers can provide plugins for specifying the type of objects to be managed and the policies they want to enforce when certain cluster operations are performed. In short, it provides an array data type for programming/managing an OpenStack cloud. The name means "forest" in Mandarin.

Who: Qi Ming Teng, PTL, is a research scientist at IBM.

What new features/enhancements/fixes are expected for the Mitaka release?
Senlin joined the OpenStack big tent during the Mitaka cycle. As a new project, the first priority for the development team is about the API specification and the stability of the service engine. Instead of delivering many fancy features to draw attention from the community, the team has chosen to focus more on making the initial release usable. The rationale behind the decision is that we can always add new features, but we will never get a chance to deprecate things we do wrong in the beginning.

In the first official release, Senlin will provide the following features:

basic management of profiles, clusters, nodes and policies
sample policies such as deletion, placement, scaling and load-balancing
binding policies to clusters
basic support to receivers (the only type would be webhook)

What contributions do you need most from the community?
Most of Senlin’s developers are from China, behind a national firewall that blocks us from accessing any social media. The team has difficulties in advertising this technology to a broader audience.

Another thing we may need help with is to raise awareness of other projects which can consume Senlin. One is Sahara, which can use Senlin to manage the VM provisionings. Another is Magnum, also can use Senlin for provisioning VMs. The advantage of using Senlin to do this is that Senlin gives the requester more knobs to control the clusters they want to manage. Using Heat for these tasks has proved to be very inconvenient.

Get involved!

Use Ask OpenStack for general questions.
For roadmap or development issues, subscribe to the OpenStack development mailing list, and use the tag [senlin]
To get code, report bugs, view blueprints, etc. check out the Senlin Launchpad page
The Senlin team holds weekly meetings on Tuesdays at 13:00 UTC in #senlin on IRC.

Cover Photo // CC BY NC