We'll soon need:
- Extra resources, mainly CPU for all of our HTTP frontend (Varnish, nginx) clusters, due to an increased use of HTTPS & IPsec
- Clusters for codfw
As HTTPS scalability is one of the quarterly goals for FY Q3 2015 and there's a lead time for procurement and site installation (plus an extra delay for esams which needs an upgrade the most, due to not having a dedicated field tech), we'll need to do this procurement very soon.
The current blocker for this is testing of our new target platform (post- T86648 & T81543) as to have more accurate figures for proceeding with procurement. Current agreement on planning suggests:
- 2nd week of January + 2-3 first weeks of February for limited performance testing
- Late February/early March for proceeding with procurement in parallel to reinstallation of the current systems
- Rest of March: setup & install of the new servers