Monitoring and graphing frameworks

Registered by Mathias Gug on 2010-04-28

Review monitoring framework.
Develop monitoring probes: one probe to rule them all (monitoring framework).
Syslog aggregation.

Blueprint information

Status:
Complete
Approver:
Jos Boumans
Priority:
Medium
Drafter:
Mathias Gug
Direction:
Approved
Assignee:
Clint Byrum
Definition:
Superseded
Series goal:
Accepted for maverick
Implementation:
Deferred
Milestone target:
None
Completed by
Mathias Gug on 2010-10-19

Related branches

Sprints

Whiteboard

Status:
Rest of work postponed indefinitely in favor of UEC monitoring.

Complexity:
maverick-alpha-3: 4

Roadmap Notes:
Timeboxed to 4 points. Evaluate implementation and review.

Work items for maverick-alpha-3:
Define cache schema (sqlite): DONE
Build schema data collection library: DONE
Write local cache collector: DONE
Write maintenance cron jobs: POSTPONED
package local cache collector: POSTPONED
Test collectd probes collection: POSTPONED

Work items:
Investigate munin probes integration with DBUS: TODO
Write munin probes integration with DBUS: TODO
Test munin probes collection: TODO
Write PoC usar: TODO

Reviewers: ttx + jib

ttx review / 20100526:
 * Spec could use some polish :)
 * Spec only mentions solving the collection issue, which I think is a good target for 10.10
 * Should move "work items" from spec to whiteboard
 * Suggested assignees: mathiaz / SpamapS
 * Estimated complexity: 6-8 (can be trimmed to 4-5 by not doing collectd and munin integration)
 * Suggested priority: 2/Medium
 * Suggested Subcycle: Iteration 1 or 2 (Alpha2 or Alpha3) -- before FF
 * Needs to be timeboxed to 4 complexity points

jib review / 20100526:
 * Prototype/verification needed for design. I fear IO bottlenecks on the sqlite implementation
    * sqlite bottleneck concern is valid, alternative of tokyocabinet should be simple to implement (SpamapS)
 * High complexity; are there milestonable deliverables in between?
 * Assignee: SpamapS - enough information here to proceed to implementation?
    * I think the spec is clear and the work items are as well. I will proceed absent any suggestion not to (SpamapS)

clint 20100719:
DBUS is removed from the spec, with collectd controlling things directly, the following work items are no longer necessary
-Define DBUS object method signature
-Write collectd write plugin over DBUS

clint 20100721:
Postponing local cache collection task for MIR of collectd and UEC monitoring scripts
-package local cache collector
-Write maintenance cron jobs
-package local cache collector
-Test collectd probes collection

(?)

Work Items