...
- Istanbul: Platform component/Handlers required to be up & running prior to DCAE Services component deployment
- Jakarta: MS deployment are uniquely handled and controlled via independent charts
Resource Optimization
Helm transformation Saving: 100% (all Platform component dependency removed)
Below table captures resource saving with DCAE transformation completed in Jakarta release
| Deployment | Istanbul | Jakarta | Overall Resource Savings | ||
|---|---|---|---|---|---|
| CPU (minutesmilliunits) | Memory (Mimebibytes) | CPU (minutesmilliunits) | Memory (Mi)mebibytes) | ||
| DCAE Bootstrap | ~48 | ~3300 | - | - | CPU utilization reduction 100% Memory utilization reduction 100% |
| DCAE Bootstrap + DCAE Microservices | ~60 | ~4500 | ~15 | ~1470 | CPU utilization reduced by ~75% Memory utilization reduced by ~75%~73% |
| DCAE Bootstrap + DCAE Microservices + DCAE MOD | ~225~85 | ~7100~6700 | ~50~40 | ~3000~3500 | CPU utilization reduced by ~75% ~63% Memory utilization reduced by ~60%~53% |
(see kubectl top stats below)
DCAE Jakarta release deployment can be done on single node/cluster (8 core and 16GB) along with other ONAP component dependencies (e.g DMAAP, AAF)
...