Services/Sync/2011Meltdown/ActionItems

Build and document model of current behaviour, expected traffic	Ops/Eng
Bring back and enforce system-review requirement for all client behaviour changes	Eng
Identify required metrics for reliability, create dashboard to compare Nightly/Aurora/Beta/Release channels	Ops
Enhanced User-Agent data required	Eng
Standing review, one week before Rapid Release cutover day, of all changes landed and all metrics from above action	Eng/Ops
Fix backoff handling for Fx8 + add automated tests	Eng	Done: bug 691612 bug 691663 bug 691988
Add backoff handling to pre-Aurora-merge extended QA validation	Eng/QA
Document and make available current system capacity numbers	Ops
Determine and implement standard for required headroom on each server	Ops
coordinate metrics-ops observation of Aurora/Beta traffic patterns	Ops/Metrics
zeus log analysis and reporting improvements	Ops
deploy TTL pruning cron	Ops
push 2.6.x server-core update	Eng/Ops
improve understanding of platform structure outside of ops	Ops
automated stack trace processing and reporting, top crasher equivalent, server-focused	Ops/Eng
Re-triage of all outstanding bugs to identify future lurking issues	Ops/Eng
90-second minimum sync interval	Eng	Done: bug 694149
Create and document engineering contact list + escalation path for all applications	Eng
Update new user activation load test, identify optimal+maximum user uptake rate per node	Ops/Eng
gradual user reassignment per node, not per db	Ops
Ensure ongoing reliability work is accounted for in planning, and deferral of maintainance	Eng/Ops	ongoing