effie

63 ms·

661.
by effie 6y ago
SATA SSDs deployments are cheaper than NVMe deployments and ofter fast enough. EDIT: you can't directly compare Samsung Pro with gaming products such as XPG GAMMIX. They are different classes of product. You wouldn't put ADATA XPG
662.
by effie 6y ago
You are right, let's hope those 600MB/s will be the case.
663.
by effie 6y ago
With 200MB/s I'm getting one week as the lower limit for rebuild time. If the drives are used heavily while rebuilding, this can go much longer.
664.
by effie 6y ago
And because large part of price-sensitive market (consumer/small business) does not care for NVMe speeds, SATA SSD aren't really obsolete. They are just older, slower tech.
665.
by effie 6y ago
More read/write heads (not necessarily more actuators). Maybe dozen heads suspended from a radially movable bridge above the platter so they can read and write simultaneously on different tracks.
666.
by effie 6y ago
Enterprise Seagates >4TB are fine now. Most problems in the past were with 3 and 4TB models. They seem to have fixed the problem and new bigger drives seem OK.
667.
by effie 6y ago
Wow, thanks for this. It seems insanely complicated. Why not just use single fat track that announces the ID number of the corresponding data track?
668.
by effie 6y ago
When you do not want or can't use an SSD for price or reliability concerns. 1) A high-throughput MX/message broker server that is set up and works next 10 years without a need of drive replacement. 2) ZFS SLOG (or any similar scra
669.
by effie 6y ago
Yes. More heads on single actuator would be more desirable, but are probably hard to make work.
670.
by effie 6y ago
URE is a read error, not a drive failure. So yes bigger disk increases chance of silent and loud UREs, but with URE rate 1e-15, for RAID1 disks up to 40TB are usable (you can redo the rebuild if it fails, or simply ignore the loss of data i
671.
by effie 6y ago
Agreed on 4TB to be a reasonable upper limit for RAID1 if URE rate is 1e-14. However, why would 1 in 20 rebuilds fail? And if it did, why would there be a data loss? The data is OK as long as the remaining disk is alive, you can retry the r
672.
by effie 6y ago
Oh you mean a home solution for few people who do not require 99.99% uptime, then yeah one could try to go that way although all such attempts I've seen (such as [1]) seem way too cumbersome, I would just build a silent PC, way simpler
673.
by effie 6y ago
Drive failure during a rebuild isn't a real concern unless 1) the rebuild takes time comparable to characteristic lifetime of the drive, which on average is >5years. Rebuild times are more in the realm of days, maybe coming to weeks
674.
by effie 6y ago
Using multiple drives is costly, it is much cheaper to consolidate if possible. Rebuild time per TB will actually slightly improve because of better throughput due to higher density and higher number of disks inside the drive, so the recove
675.
by effie 6y ago
Why? It is cheaper to run redundant disks and power supplies with single mainboard than to run two mainboards.
676.
by effie 6y ago
Interesting. What do you think about Seagate's claim that dual actuator decreases their costs because it takes less time to test the drive? Is that the real reason for dual actuator?
677.
by effie 6y ago
Exactly.
678.
by effie 6y ago
It is a pointless comparison, both methods contribution to global CO2 emissions is absurdly minuscule.
679.
by effie 6y ago
>waste we still have no real way to reliably deal with What is your idea of a "real way to reliably deal with waste"? Waste in general is stored or burned all over the world. Nuclear waste is best stored, at NPPs or at dedicate
680.
by effie 6y ago
The article does not show any reason to doubt safety of nuclear plants. Increased leukemia incidence in proximity to NPPs is interesting and should be further studied, but there are different possible reasons for it. It can be a societal ef
681.
by effie 6y ago
> In Qubes, you typically don't care if your untrusted application compromises the (untrusted) virtual machine. For that to be true, one would have to run each application in a dedicated VM. A great idea, but Qubes does not recommen
682.
by effie 6y ago
Qubes provides a very specific product. Whether it helps to "protect journalists and others who would greatly benefit from enhanced security" depends on user level of knowledge and the ways he uses the computer. If the user does n
683.
by effie 6y ago
Notice how that argument starts with the conclusion "In Qubes VMs there is no point in isolating the root account" and proceeds to rationalize it, very poorly, by using some cherry picked examples where that may be true. OK, an
684.
by effie 6y ago
> really is an indictment of mainly Windows but also desktop Linux. I think it is more about the market size. Only a small part of population needs or wants PCs at all, so the base for marketing the high end is limited. But almost everyb
685.
by effie 6y ago
Qubes is very "opinionated" about how things work and provides a functional but very tightly-knit product that cannot be easily modified to suit your own needs. You have to accept their choices like a good nontechnical user does -
686.
by effie 6y ago
"The Qubes people" have a product to develop and maintain. They aren't the single highest authority on secure desktop setups. Security isn't a black or white issue. There are levels of security. Many tech people want som
687.
by effie 6y ago
> I am yet to find a lucid and intuitive explanation of why we need both E and D for the electric field and B and H for the magnetic field. E alone cannot describe fully state of dielectric, and B alone cannot fully describe state of mag
688.
by effie 6y ago
Where is big money to be made, in laptops and desktops, or in overpriced phones?
689.
by effie 6y ago
> Intel allows virtual GPUs for virtualization in QEMU which AMD won't do. What is the level of capabilities we can expect from this? How many simultaneous VMs can use the Intel GPU?
690.
by effie 6y ago
> That section mentions a wide range of policy implications—several paragraphs that you didn't bother quoting—without advocating for anything in particular beyond reducing harm. Yes but that is still propagating the unsubstantiated

More ›