starting...
loading...
Markov chain of the selected scheme: states count degraded drives, rates in 1/hour
running...
The science
Same model and the very same module as the 3D page. Failure model. A drive with annual failure probability AFR fails at a constant rate λ = −ln(1 − AFR) / 8766 h. A failed drive is rebuilt in an exponentially distributed time with mean MTTR. Mean time to data loss (MTTDL) comes from a continuous-time Markov chain whose states count degraded drives (or degraded mirror pairs): solving (−Q) t = 1 gives the mean time to the absorbing "data lost" state, and the loss probability by year t is read from exp(Q t). The chain diagram under the graphs draws exactly that matrix Q.
RAID 5: MTTDL ≈ MTBF² / (N (N−1) MTTR) RAID 6: ≈ 2 MTBF³ / (N (N−1)(N−2) MTTR²) RAID 10: ≈ MTBF² / (N MTTR)
Read errors. Rebuilding RAID 5 reads every surviving drive in full. With an unrecoverable read error rate b per bit, the chance of at least one error is 1 − exp(−b · bits read), which is why large single-parity arrays of consumer drives are discouraged. RAID 6 can still repair one such error.
Why more parity helps so much. Each extra layer multiplies MTTDL by roughly MTBF / (N MTTR). Shortening the rebuild window helps quadratically for RAID 5 and RAID 10 and cubically for RAID 6, which is why a hot spare matters.
Checks. The Markov results are compared with the textbook closed forms in their valid regime, and with a seeded event-by-event Monte Carlo of 3000 rack lives in an accelerated stress case (the same random-drive code that drives the rack animation). The last check confirms that the chance of loss within one MTTDL is 1 − 1/e. The timeline records drive states at sub-steps of at most a third of the MTTR, so very short rebuilds at extreme time-lapse speed may be drawn slightly coarse; the statistics themselves use the exact event times.
Simplifications. Drive failures are independent and constant-rate: real drives show infant mortality and wear-out, batches fail together, and heat or a power surge hits many drives at once. The rebuild time is the same for every drive and does not slow down under load. Controller, software and human errors, and RAID write holes, are not modelled. RAID is not a backup: it protects against drive failure, not deletion, ransomware or fire, so an off-site copy (the vault) is still needed. Sprites are Blender 5.2 renders of the 3D page's models.