C1 (3 points)
https://knobit.org/id/task/dso-recovery-service-levels/C1
Also as Turtle · JSON-LD · RDF/XML
| dcterms:license | https://creativecommons.org/licenses/by/4.0/ |
|---|---|
| rdfs:label | C1 (3 points) en |
| schema:hasPart | dso-recovery-service-levels C1a: SL-101 recovers along the OL-101 -> OL-102 chain, past CS-11's tolerance |
| schema:hasPart | dso-recovery-service-levels C1b: OL-104's synchronous replica is no copy against corruption |
| schema:hasPart | dso-recovery-service-levels C1c: OL-107's partial restore test does not prove the RTO |
| schema:identifier | C1 |
| schema:position | 1 |
| schema:text | Recovery commitments against how recovery actually works, 1 each. (a) SL-101: instant payments are back only when the whole chain is: the payment hub database (OL-101) needs PT1H45M and the SEPA Inst gateway (OL-102) can start only after it, PT30M more, so the service recovers in PT2H15M in both scenarios (core banking, OL-103, recovers in parallel and does not change this) - over CS-11's PT2H tolerance, though each OLA looks fine alone. Reported, with a remedy proposed for the owners (gateway recovering in parallel or kept warm, faster database recovery) or the gap raised with the Head of Payments; the tolerance is not touched. (b) SL-102 and OL-104: a synchronous replica is not a backup - it copies corruption and ransomware to the second site at once, and OL-104 has no point-in-time copy at all, so the corruption RPO of PT0S (and RTO of PT15M) is not real. Immutable point-in-time backups outside the primary's failure domain proposed for OL-104 (cards platform lead to agree), with the corruption RPO taken from their interval or left null and raised; the site-loss PT0S at 25 km stays. (c) SL-103 and OL-107: the June restore test restored 1.5 TB of the 12 TB channel store, so it does not prove the PT2H corruption RTO: a full-scope restore test planned (no result invented), and the risk raised that the full restore takes longer - scaled linearly the test suggests about PT5H20M, beyond CS-13's PT4H. en |