Some NVMe drive temperature things from my drives

I said some things on Twitter about thetemperatures of the NVMe drives in my home and work machines (asreported by the Linux kernel), so I'm going to write down more onthat here. I don't know if it means anything, but it's at leastsome data points.

Both my work desktop and my homedesktop now have a pair of NVMe drivesin them, all of which report the nominal drive temperature to theLinux kernel. The machines are built using the same case


NVMe drives and the case of opaque bandwidth limits

I tweeted :

Given that my home desktop's two M.2 NVMe slots seem to share a singlePCIe x4 uplink in the end, it's probably not worth losing two (unused)SATA ports to bring the second NVMe drive up from x2 to x4 since I'mmirroring them anyway.

My current home desktop has an AsusPRIME Z370-A motherboard. This motherboard has two M.2 NVMe slots,one that's always at x4 PCIe lanes ,


Linux's vm.admin_reserve_kbytes sysctl is both not big enough and not sufficient

We enable Linux's strict overcommit on someof our servers (mostly compute servers). Every so often people runbig enough programs that they run the machine out of memory, andsome of the time when this happens we get various plaintive reportsfrom cron and other things that periodic system processes havefailed with out of memory errors. The Linux kernel has a sysctlthat's supposed to help with this, vm.admin_reserve_kbytes (documented in vm.txt ), butin practice we'


A bit on what Unix system pre-boot environments used to look like

Unix was first implemented on general purpose computers that hadn'tbeen specifically designed for it, such as the PDP-11 and the DECVax. These machines could have intricate start up procedures ( also ) and in any casetheir pre-boot environment wasn't designed for Unix specifically.This changed in the early 1980s when computers both got more complexand began to be designed specifically for Unix, such as the Sun-1 . These Unix computers, designedand built by Unix vendors


Most computer standards broadly require good faith implementations

One of the quietly unsaid things about computer standards, both formaland relatively informal, is that they were generally not created todefend us against deliberately perverse implementors with maliciousintent. Most useful computer standards aren't written so tightlyand narrowly that they block out useless but conforming behavior(or perverse but conforming); they assume that people following thestandards are actively attempting to deliver a useful result. Theimplementors may misunderstand things or be sloppy, but they aren'ttrying to be evil.

This is broadly not


My views so far on bookmarklets versus addons in Firefox

A while back I saw a message in the fediverse advocating forbookmarklets ( via ):

90% of browser extensions could be bookmarklets, and they’re 100%better for privacy. Bookmarklets should make a comeback.

* Opt in to an action
• No snooping on pages in the background
• Universally supported in browsers

I miss bookmarklets.

I'm a heavy user of Firefox addons , and I'vealso recently been using bookmarklets in the titlebar for dealing with fixed positionpage elements and making pages


On servers maybe moving to M.2 NVMe drives for their system drives

We 've been looking into gettingsome new servers (partly because a number of our existing DellR210 IIs are starting to fail).Although we haven't run into this yet ourselves, one of the thingswe've heard in the process of this investigation is that variouslines of basic servers are trying to move to M.2 NVMe system disksinstead of 3.5" or 2.5" disks. People are generally unhappy aboutthis, for a number of reasons including that


Unfortunately, damaged ZFS filesystems can be more or less unrepairable

An unfortunate piece of ZFS news of the time interval is that Ubuntu21.10 shipped with a serious ZFS bug that created corrupted ZFSfilesystems (see the 21.10 release notes ; via ).This sort of ZFS bug happens from time to time and has likely happenedas far back as Solaris ZFS, and there are two unfortunate aspects of them.

(For an example of Solaris ZFS corruption, Solaris ZFS couldwrite ACL data that was bad in a way that it ignored but modern ZFS


Prometheus will make persistent connections to agents (scrape targets)

The other day, I discovered something new to me about Prometheus :

Today I learned that the Prometheus server will keep persistentTCP/HTTP connections open to clients it's pulling data from if you'repulling data frequently enough (eg, every 15 seconds). This can matterif you add firewall rules that will block new connections to things.

Julien Pivotto added an important additional note :

It also [h]as an impact if you use DNS because we will only resolvewhen we make new connections

I


The long term relative prices of M.2 NVMe drives and 2.5" SSDs

For reasons outside the scope of this entry, I recently found myselfwondering if in the end, M.2 NVMe drives will wind up being lessexpensive at moderate capacities than 2.5" SSDs of the same capacity(regardless of the 2.5" interface interface involved, which mightbe SATA, SAS, or U.2 NVMe). I reflexively think of M.2 NVMe drivesas a better, high end product that is and always will be moreexpensive than ordinary 2.5" SATA