All Posts by Date or last 15, 30, 90 or 180 days.
also by Lloyd: diglloyd.com photography and WindInMyFace.com
Thank you for purchasing through links and ads on this site.
OWC / MacSales.com...
diglloyd Deal Finder...
Buy other stuff at Amazon.com...
Upgrade the memory of your 2019 iMac up to 128GB

Testing Hard Drive Consistency for a RAID

diglloydTools

Purchase diglloydTools...

Among its many features, diglloydTools DiskTester offers a run-area-test command that can be used for relatively fast characterization of drive performance across its capacity. The results can be graphed to get a good idea of the drive performance.

For example, suppose a high performance RAID is envisioned: one laggard can cut the performance, since the slowest drive determines the peak speed.

As shown below, 8 samples were tested, and one of those eight is a significant laggard; this slower drive is best set aside as a cold spare or backup drive.

While DiskTester can test any number of drives simultaneously (via command line), testing should take care not to saturate the bandwidth; for example four fast hard drives can demand almost all of the Thunderbolt v1 bandwidth, and other devices on the bus could suck up bandwidth, disturbing the results.

Click for larger graph.

diglloydTools DiskTester: performance across volume capacity for eight drives
disktester run-area-test --iterations 5 --test-size 4G --delta-percent 5

DiskTester fill-volume — more detailed evaluation

For production RAID, MPG strongly recommends testing each drive individually with DiskTester fill-volume, to weed out any laggards and/or any drives with performance oddities. The fill-volume test takes a long time, since it has to write and read the entire drive. That’s 16TB for an 8TB drive, so at an average hard drive speed across the capacity of ~150 MB/sec, the test cycle will take about 29 hours.

The DiskTester fill-volume writes 1000 files that fill the capacity of the drive, and then reads them back, generating output data that can be pasted into the supplied spreadsheet for a graph that shows performance in detail across the entire capacity. Any glitches such as ranges of bad blocks, drive behaviors such as pauses for recalibration, etc. Another good use is for comparing drives; see 4/5/6TB Hard Drives: Higher Capacity Boosts Real-World Performance. As an added benefit, actually writing the entire drive can flush out problem drives that have a range of bad blocks (most hard drives ship with untested media!).

Hard drive performance vs capacity utilization over 4TB partition (4/5/6TB hard drives)

It is possibly to test any number of drives simultaneously with DiskTester, one per Terminal window, but beware of bus bandwidth limitations, particularly with SSDs, or too many drives will compete and all will be slowed by the others. For example, Thunderbolt 2 maxes-out at about 1350 MB/sec and all Macs except the Mac Pro have only a single Thunderbolt bus.

Shown below is DiskTester fill-volume being run in five Terminal windows on five HGST Ultrastar He8 hard drives. The HGST He8 enterprise-grade hard drive is unusually consistent sample-to-sample as can be seen by the "File #" and “ClockAvg” (MB/sec) column for each. The fill-volume test takes about 14 hours to write the data, and 14 hours to read it back, so it is a long running test.

Running diglloydTools disktester fill-volume on five HGST Ultrastar He8 8TB hard drives
disktester fill-volume

Some of the capabilities in diglloydTools

Aside from testing hard drive or SSD or RAID performance and reliability with DiskTester, data integrity with IntegrityChecker is a must-have workflow tool for anyone with important data:

diglloyd.com | Terms of Use | PRIVACY POLICY
Contact | About Lloyd Chambers | Consulting | Photo Tours
Mailing Lists | RSS Feeds | Twitter
Copyright © 2019 diglloyd Inc, all rights reserved.
Display info: __RETINA_INFO_STATUS__