For a parity array, no. Day-to-day use is usually fine, and that is what misleads people, the problem appears during a rebuild or a resilver, which is precisely the moment your data has no redundancy protecting it.
Shingled recording overlaps data tracks like roof tiles to raise density. Reading is unaffected. Writing is affected when the drive has to modify data underneath an existing track, because the tracks laid over it may have to be rewritten too. That penalty is invisible for sequential work and severe for sustained random writing, which is what a rebuild or a resilver actually is.
In normal operation a shingled drive hides the penalty well. It keeps a conventionally-recorded staging area on the platter, absorbs incoming writes there, and reorganises during idle time. For a household NAS writing a few gigabytes a day and idle most of the night, that mechanism works and you would never know.
A rebuild removes both conditions at once. The array writes continuously for hours or days with no idle periods, and the volume far exceeds any staging area. The drive falls back to its slow path and stays there. Rebuilds that would take hours can take very much longer, and some controllers time the drive out and drop it from the array, which turns a degraded array into a failed one.
That is the whole risk in one sentence: shingled recording is fine until the moment you most need it not to be.
It would be dishonest to leave this as a blanket warning, because there are real cases where shingled drives are the correct economic choice.
A single-drive backup target with no parity is one. There is no rebuild, no resilver, and the workload is long sequential writes, which is what these drives handle well. A cold archive written once and read occasionally is another. A media library where new files are added whole and never modified is a third.
JBOD and unRAID-style arrays that do not stripe with parity sit in between: there is no traditional rebuild across all drives, so the worst case is milder, though a parity-protected unRAID array does still rewrite a full drive during recovery.
The common thread is the absence of sustained random writing across the whole drive. Where that is genuinely absent, the cost saving is real and worth taking.
Read the part number rather than the product name. The name tells you the family and the marketing position; the part number tells you the drive. Our resolver takes a pasted model number and answers from the manufacturer's own document.
The trap to know about is that a manufacturer can change the recording technology within a product line without changing the name. Seagate's 1TB BarraCuda is the clearest case: the ST1000DM010 is conventional, the ST1000DM014 that replaced it is shingled, and both are sold as a BarraCuda 1TB.
If a drive is already in your array and you have discovered it is shingled, do not panic and do not necessarily replace it immediately. Make sure your backups are current, avoid unnecessary rebuild-triggering operations, and replace it at the next natural opportunity rather than forcing a rebuild today to remove it.
| Drive | Capacity | Recording | $/TB | Price | |
|---|---|---|---|---|---|
| 695842-001 4TB SAS 6G | 4TB | CMR | $14.00 | $56 | Buy |
| HUS726040ALS210 4TB SAS 12G | 4TB | CMR | $18.68 | $75 | Buy |
| 695842-001 4TB SAS 6G LFF | 4TB | CMR | $20.56 | $82 | Buy |
| 012GYY 4TB SAS 6G Renewed | 4TB | CMR | $21.25 | $85 | Buy |
| DRMYH Compellent 4TB NL SAS Renewed | 4TB | CMR | $22.50 | $90 | Buy |
It will be recognised and will work day to day. The failure case is the rebuild: continuous writing for hours with no idle time is the pattern shingled recording handles worst, and some controllers will time the drive out and drop it, converting a degraded array into a failed one.
Yes, in specific shapes. A single-drive backup target, a cold archive, or a media library where files are written whole and never modified are all fine, because none of them involve sustained random writing or a parity rebuild. The saving in those roles is real.
By its part number, not its product name. Manufacturers change recording technology within a line without renaming the product, so the name is not reliable. Paste the part number into our resolver and it will answer from the manufacturer's own document, with the source linked.
Do not force a rebuild today just to remove it. Verify your backups are current, avoid operations that trigger a full resilver, and replace it at the next natural opportunity. The risk is concentrated in rebuild events, so the worst response is to create one unnecessarily.