All of lore.kernel.org
 help / color / mirror / Atom feed
From: Roman Mamedov <rm@romanrm.net>
To: Gandalf Corvotempesta <gandalf.corvotempesta@gmail.com>
Cc: Wols Lists <antlists@youngman.org.uk>,
	Linux RAID Mailing List <linux-raid@vger.kernel.org>
Subject: Re: SMR or SSD disks?
Date: Mon, 27 Nov 2023 03:52:37 +0500	[thread overview]
Message-ID: <20231127035237.3e7d78da@nvm> (raw)
In-Reply-To: <CAJH6TXiKLitvausMWgwZ_kNQbvp7sDA4AfaLvO2M54E7qLq8cg@mail.gmail.com>

On Sun, 26 Nov 2023 23:22:51 +0100
Gandalf Corvotempesta <gandalf.corvotempesta@gmail.com> wrote:

> Il giorno dom 26 nov 2023 alle ore 13:22 Wols Lists
> <antlists@youngman.org.uk> ha scritto:
> > If you look at what SMR is, it's only relevant to spinning rust. It
> > relies on the fact that a read head can be much smaller than a write
> > head, so provided you shingle your writes (hence the name), you can
> > over-write half the previous track (so saving space) without rendering
> > the data unreadable.
> 
> Thank you all.
> That's what i've thought but better stay safe than sorry so i've asked.
> 
> In other words WD Red SSD are safe for a RAID, there is no need to change them
> (as mostly new) in both array (the grow was finished 1 hour ago: 2 WD
> Gold SATA 3.5 plus 1 WD RED SSD)
> 
> Slowly, i'll replace all spinning disks with WD Red SSD
> I'm not a fan of WD, but 2TB disks able to replace a 2TB HDD are very
> very rare (the 1.92TB, much more common, can't replace a 2TB disk)

There are three grades of capacity that you can get:
1) ~1920 000 000 000 bytes
2) ~2000 000 000 000 bytes
3) ~2048 000 000 000 bytes

Nobody is marketing the 1st variant as "2TB", you will find "1920G" on the
packaging and datasheets instead.

2nd one should be able to replace an HDD, unless there's some smaller
discrepancy in the sizes near the 2 billion byte mark between HDDs and SSDs. A
reason to use smaller-than-whole-disk partitions for your RAID.

3rd is not a common sight in SATA SSDs (but still happens), and is nearly
universal in NVMe. Of about ten "2 TB" NVMe models I've tested, only one had
2000, *all* others were 2048. The 2000 one was Netac NV7000.

-- 
With respect,
Roman

  reply	other threads:[~2023-11-26 22:52 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-11-26 10:18 SMR or SSD disks? Gandalf Corvotempesta
2023-11-26 10:37 ` Andy Smith
2023-11-26 11:55   ` Hannes Reinecke
2023-11-26 12:22     ` Wols Lists
2023-11-26 22:22       ` Gandalf Corvotempesta
2023-11-26 22:52         ` Roman Mamedov [this message]
     [not found]           ` <CAJH6TXi2xc9o95qp_UfyyqSW=H2ssK-ZBvnfP3vpw89umoqD5A@mail.gmail.com>
2023-11-27  8:11             ` Gandalf Corvotempesta
2023-11-27  9:15               ` Reindl Harald
2023-11-27 12:35                 ` Gandalf Corvotempesta
2023-11-27 12:36                   ` Gandalf Corvotempesta
2023-11-27 17:23                     ` Reindl Harald
2023-11-27 18:56                       ` Gandalf Corvotempesta
2023-11-27 17:18                   ` Reindl Harald
2023-11-26 21:17     ` Peter Grandi

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20231127035237.3e7d78da@nvm \
    --to=rm@romanrm.net \
    --cc=antlists@youngman.org.uk \
    --cc=gandalf.corvotempesta@gmail.com \
    --cc=linux-raid@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.