I’m looking at a few nvme options to run a node :
#1 SanDisk 内蔵 M.2-2280 SSD/Extreme Pro 1TB / PCIe Gen3 NVMe / 5年保証/ SDSSDXPM2-1T00-G25
Read 3,400 MB/s
Write 2,800 MB/s
500K IOPS Read
400K IOPS Write
Reliability 300 TBW
—
#2 SAMSUNG サムスン 970 PRO MZ-V7P1T0YO3 [NVMe SSD 970 PRO M.2 1TB バックアップSWダウンロード付] MZ-V7P1T0B/IT同様
Read 3,500 MB/s
Write 2,700 MB/s
? IOPS Read
? IOPS Write
Reliability ?
—
General question about measuring the reliability of SSDs and how much can expect their life expectancy to be shortened under the load of running a node ?
Total terabytes written over time (TBW)
Drive writes per day (DWPD)
Mean Time Between Failures (MTBF)
The requirements page :
Storage 1: 1TB High Speed Enterprise NVME (PCI) SSD
Recommended speed: 500,000 /500,000 IOPS
Recommended Reliability: 1.5 million hours MTBF
Example: Samsung 970 PRO SSD 1TB - M.2 NVMe
Note: SSD reliability is not as important for the BetaNet when there will not be a large number of transitions. It is likely that lower endurance SSDs will be fine but will need to be replaced when it comes time for MainNet. If an SSD fails, it could take substantial time before a node is able to get back online.
—
cheers,