From owner-freebsd-questions@freebsd.org Wed Oct 14 18:22:24 2020 Return-Path: Delivered-To: freebsd-questions@mailman.nyi.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2610:1c1:1:606c::19:1]) by mailman.nyi.freebsd.org (Postfix) with ESMTP id 2B19543FF66 for ; Wed, 14 Oct 2020 18:22:24 +0000 (UTC) (envelope-from ralf-mardorf@riseup.net) Received: from mx1.riseup.net (mx1.riseup.net [198.252.153.129]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client CN "*.riseup.net", Issuer "Sectigo RSA Domain Validation Secure Server CA" (verified OK)) by mx1.freebsd.org (Postfix) with ESMTPS id 4CBLNv1gWhz4Kbn for ; Wed, 14 Oct 2020 18:22:22 +0000 (UTC) (envelope-from ralf-mardorf@riseup.net) Received: from capuchin.riseup.net (capuchin-pn.riseup.net [10.0.1.176]) (using TLSv1 with cipher ECDHE-RSA-AES256-SHA (256/256 bits)) (Client CN "*.riseup.net", Issuer "Sectigo RSA Domain Validation Secure Server CA" (not verified)) by mx1.riseup.net (Postfix) with ESMTPS id 4CBLNj1yfnzDsZj for ; Wed, 14 Oct 2020 11:22:13 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=riseup.net; s=squak; t=1602699733; bh=nvq5UDTsSOTH8isxmrBEFdZ0dyNJfys/5uEI7/ffv68=; h=Date:From:To:Subject:In-Reply-To:References:From; b=Q/W8WfrWAgB4iOOnEq96ZkBxEsUA2VZZs0JhBZo7Dq9O4vihe4hGh1/mD/M+CnNe1 urr3if1XGRurIAjirDZ0GExLrVzZQ8iRC+amSGVyaoJrvzuPHifWGyWNrjumF5P5HI DF2JTYqnnsP/0rqneusdcRzV9o0e8zAuTfcbxKLI= X-Riseup-User-ID: F9A2F3557620BEEE51EE69811187BA7B1FEDCCE25146A2BCB11902ACA6D0CD19 Received: from [127.0.0.1] (localhost [127.0.0.1]) by capuchin.riseup.net (Postfix) with ESMTPSA id 4CBLNh4R8wz8trr for ; Wed, 14 Oct 2020 11:22:12 -0700 (PDT) Date: Wed, 14 Oct 2020 20:22:06 +0200 From: Ralf Mardorf To: freebsd-questions@freebsd.org Subject: Re: A couple of questions about SSDs Message-ID: <20201014202206.7c7886d0@archlinux> In-Reply-To: <20201014174749.6df7572a.freebsd@edvax.de> References: <20201014121442.662e71c4@archlinux> <20201014174749.6df7572a.freebsd@edvax.de> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable X-Rspamd-Queue-Id: 4CBLNv1gWhz4Kbn X-Spamd-Bar: ---- Authentication-Results: mx1.freebsd.org; dkim=pass header.d=riseup.net header.s=squak header.b=Q/W8WfrW; dmarc=pass (policy=none) header.from=riseup.net; spf=pass (mx1.freebsd.org: domain of ralf-mardorf@riseup.net designates 198.252.153.129 as permitted sender) smtp.mailfrom=ralf-mardorf@riseup.net X-Spamd-Result: default: False [-4.03 / 15.00]; RCVD_VIA_SMTP_AUTH(0.00)[]; ARC_NA(0.00)[]; R_DKIM_ALLOW(-0.20)[riseup.net:s=squak]; NEURAL_HAM_MEDIUM(-0.99)[-0.986]; FROM_HAS_DN(0.00)[]; RWL_MAILSPIKE_GOOD(0.00)[198.252.153.129:from]; TO_MATCH_ENVRCPT_ALL(0.00)[]; R_SPF_ALLOW(-0.20)[+mx]; MIME_GOOD(-0.10)[text/plain]; TO_DN_NONE(0.00)[]; PREVIOUSLY_DELIVERED(0.00)[freebsd-questions@freebsd.org]; RCPT_COUNT_ONE(0.00)[1]; DWL_DNSWL_LOW(-1.00)[riseup.net:dkim]; RCVD_COUNT_THREE(0.00)[3]; DKIM_TRACE(0.00)[riseup.net:+]; DMARC_POLICY_ALLOW(-0.50)[riseup.net,none]; NEURAL_HAM_SHORT(-0.42)[-0.424]; NEURAL_HAM_LONG(-1.02)[-1.016]; FROM_EQ_ENVFROM(0.00)[]; MIME_TRACE(0.00)[0:+]; MID_RHS_NOT_FQDN(0.50)[]; ASN(0.00)[asn:16652, ipnet:198.252.153.0/24, country:US]; RCVD_TLS_ALL(0.00)[]; MAILMAN_DEST(0.00)[freebsd-questions]; RCVD_IN_DNSWL_LOW(-0.10)[198.252.153.129:from] X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.33 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 14 Oct 2020 18:22:24 -0000 On Wed, 14 Oct 2020 17:47:49 +0200, Polytropon wrote: >My guess (and it's really just a guess due to my personal lack >of any important data and experience) is that the SSD will work >until it is replaced by a bigger one, or the whole system it is >installed in will be replaced with a faster one. So there is >nothing to worry about. If you _really_ want to worry about >something, ask the manufacturer if they made the firmware brick >the whole thing when a certain write count is reached so you >cannot even _read_ your data anymore. :-) In my desktop PC are 5 SSDs. Four are connected to the mobos SATA 3 connectors and one is connected to a SATA 2 connector. Four are 223.57 GiB sized and one is 447.13 GiB sized. I'm to lazy to check how old each of them is, but IIRC the oldest is around 3=C2=BD years old and the vendor's software mentions that the "health" is at 64%, the system drive is one of the newer SSDs, maybe around 2 years old, "health" 57%. In my experiences HDDs last for around 2 years, if you turn the computer on and off very often and for around 7 years, if the computer runs more or less 24/7. It depends on several factors ;). I don't know for how long external backup and archive HDDs do last. Probably for way longer than 7 years. However, my guess is, due to a lack of experiences with SSDs, that they last for as long as HDDs do last. However, I suspect I will replace one by another with more sized SSDs already before they fail. I became a digital photographer a few weeks ago and noticed, that I was mistaken a while ago, when we talked about storage space. Nowadays I tend to make one or two photos more, than I've done, when I needed to pay for 35 mm film ;). Due to a lack of money, I put together two new 2 external 2 TB HDDs with USB enclosures, to reorganize my current internal SSDs and my current external backup and archive HDDs. Btw. I've done this today. As soon as I've got enough money again, I will get also more external storage space. Maybe I'll continue using USB (/eSATA) enclosures, maybe I'll get a swap bay. I've got no experiences with NAS. It seems to be too expensive. However, a starting point would be to replace my 1 and 2 TB external HDDs with 4 TB HDDs. The used enclosures are suitable for HDDs up to 4 TB. >As long as the light switch doesn't suddenly shout "I've been >used 1500 times now, I'm kaputt, please buy a new one!" everything >should be okay within the limits of reality. If my SSDs should do this after 4 or 7 years, I could accept it. Btw. my iPad 2 is way older than 3 years and the internal 32 TB drive was used to it's limits, but still works. My new iPad has a build in 1 TB drive for good reasons. >For the case of using lots of "throwaway files" (i. e. stuff you >only need as files during ports builds), you can always use a >RAM disk, if your system has enough RAM. And swap space that you >don't write to doesn't add any wear. FreeBSD might work better, maybe the Linux of other users does, too. If I exceed tmpfs size with my Linux, when building packages, swap isn't used. Build aborts with a "no space left on device" error. >> What exactly makes you think, that SSDs need gentle treatment? =20 > >It's probably the limit on write cycles, but I'm not sure how >this compares to general lifetime calculations compared to >regular hard disks... Neither do I, but it seems likely that they last as long, as HDDs do. Most file systems are ext4 (on the old HDDs it probably was ext3), journaling enabled, some use relatime, some use noatime. The noatime option isn't used to reduce write cycles of a SSD, it's a leftover from the times when my computers used HDDs for audio production and should help to achieve better performance.