November 2021 - Talk - lists.gtalug.org

war story: power failure makes CentOS system unbootable
by D. Hugh Redelmeier 30 Sep '22

30 Sep '22

tl;dr / spoiler: change fstab so that /boot/efi is not automatically mounted. See the recommendation at the end of this message. My gateway computer is a little PC running CentOS. It does not come back after a power failure. The reason (as best I can tell) is interesting and I think that I have a fix. My system is UEFI. It boots from a UEFI system partition, known as /boot/efi to Linux. If this gets corrupted, it won't boot. It is a VFAT partition. On my gateway (a Zotac Zbox RI321, a cute little box with two ethernet ports), the UEFI firmware apparently won't boot if the dirty bit is set in the system partition. CentOS normally runs with /boot/efi mounted. So when the system isn't shut down cleanly, the dirty bit will be on. Consequence: the system is not going to boot after a power failure. Odd observation 1: if I put a live Fedora 25 USB stick in the system and try to boot in this otherwise-unbootable state, CentOS boots from the main disk. So this really looks like a firmware bug/quirk. Odd observation 2: fsck doesn't seem to automatically fix the system partition. Once CentOS 7 booted, I dismounted /boot/efi and did an fsck on it. $ sudo fsck /dev/sda1 fsck from util-linux 2.23.2 fsck.fat 3.0.20 (12 Jun 2013) 0x25: Dirty bit is set. Fs was not properly unmounted and some data may be corrupt. 1) Remove dirty bit 2) No action ? 1 Leaving filesystem unchanged. /dev/sda1: 16 files, 2420/51145 clusters Googling got me to <https://www.centos.org/forums/viewtopic.php?t=50917> In particular, this advice seemed quite good: ==== recommendation ==== Two other changes I recommend for UEFI systems, to each OS's /etc/fstab. - For the /boot/efi mountpoint, add the mount options x-systemd.automount,noauto - Change fs_passno (last column) to a 1 or 2; the canonical fstab instructions suggest 2, but systemd treats 1 and 2 as the same. The first change means the EFI System partition will not be automatically read-write mounted at boot time; it's a bad idea this is the default because it puts the ESP at risk especially if ther are crashes, FAT has no journal and will therefore always be marked dirty in such a case; no other UEFI OS mounts the ESP by default. Second, if anything tries to access /boot/efi (read or write), systemd will automatically mount it, and because of the fs_passno 1 or 2, it will fsck it first and that fixes and clears the dirty bit in case it's set. Right now without these changes, it's just a matter of having the right number and bad timing of crashes to render the EFI System partition corrupt.

3 9

Booting linux from nvme disk?
by Peter King 10 Dec '21

10 Dec '21

Anyone had success with getting linux to boot from an nvme disk? I've been fighting with UEFI and Arch Linux all day now, trying to get a new computer to even recognize the nvme disk as a boot device. (Last time I encountered this problem I gave up and installed an ordinary HD to boot from.) For what it's worth, I have disabled Secure Boot, reformatted the nvme disk to have a new EFI partition without Windows, a swap partition, and a root partition; I'm trying to use efibootmgr (so no loading of a further bootloader), and, as far as I can tell, identified all the right partitions by device name or PARTUUID. Still no go. I even tried adding the parameter nvme_load=YES into the "root" part of the efibootmgr, and also adding nvme and vmd as modules in mkinitcpio.conf, all without any success. If necessary I'll just punt and install a regular HD to boot from, but that rather takes away from having an nvme disk in the first place. (Over the years I've learning to approach installing Linux with fear and loathing, with almost all the problems being with the bootloader -- from LILO through GRUB and GRUB2 now down to UEFI.) Any suggestions welcome! I didn't want to spend my weekend doing this. -- Peter King peter.king(a)utoronto.ca Department of Philosophy 170 St. George Street #521 The University of Toronto (416)-946-3170 ofc Toronto, ON M5R 2M8 CANADA http://individual.utoronto.ca/pking/ ========================================================================= GPG keyID 0x7587EC42 (2B14 A355 46BC 2A16 D0BC 36F5 1FE6 D32A 7587 EC42) gpg --keyserver pgp.mit.edu --recv-keys 7587EC42

10 14

a solved problem unsolved itself: WordPress, MySQL, UTF-8
by Stewart C. Russell 02 Dec '21

02 Dec '21

I have been running a WordPress blog hosted on a Linux-based shared host since WordPress became a thing. It has worked quite well from about 2004 up until a few weeks ago. Sadly, *something* recently decided my database encoding was wrong. And that something decided to "fix" it. It certainly "fixed it", but not in any way I could want. It also did the same for Catherine's blog. I know I didn't change any part of the config chain. As far as I can see: * the MySQL database still thinks the text is encoded in UTF-8; * Wordpress thinks the data is in UTF-8; * the web server is serving UTF-8. (NB: there's going to be some UTF-8 and hex chars in this message.) A typical post which shows this problem is https://scruss.com/blog/2016/02/27/t%c9%92k-b%c9%92ks-a-tiny-hardware-speec… When I should be seeing something like: [tɒk bɒks] 5b74 c992 6b20 62c9 926b 735d I'm seeing this in the page served up (and in the db text itself): [tÉ’k bÉ’ks] 5b74 c389 e280 996b 2062 c389 e280 996b 735d So the phonetic character U+0252 has been mangled into U+00C9 + U+2019. Every UTF-8 character seems to be affected this way. I wasn't expecting to wake up to a UTF-8 encoding problem this decade. There are a raft of "how to fix WP encoding issues" pages that show up in web searches, but the newest of them is from 2008 or so. I'm pretty much resigned to going through 16+ years of posts fixing this, but can mangled UTF-8 be recovered without rekeying? cheers, Stewart

6 11

Index Exchange is hiring (a lot), and I can recommend it
by Dave Collier-Brown 01 Dec '21

01 Dec '21

We're an advertising exchange that does real time ad auctions in 120 milliseconds, growing hugely even through COVID, financially successful and moving steadily toward IPO in a few years. We're interested in Go programmers, sysadmins with a talent for automation, and data / machine-learning engineers. Plus co-ops. There are roughly 60 engineering position we want to fill, nominally in downtown Kitchener-Waterloo, Toronto and Montreal. In fact, we've mostly been working from home world-wide, and turned our old North Toronto office into a data centre. We do have a building at King and Spadina for folks who prefer to have a desk (which includes me), but it's optional. This is the best job I've had since Sun Microsystems: come join me. --dave -- David Collier-Brown, | Always do right. This will gratify System Programmer and Author. | some people and astonish the rest dave.collier-brown(a)indexexchange.com<mailto:dave.collier-brown@indexexchange.com> | -- Mark Twain davecb(a)spamcop.net<mailto:davecb@spamcop.net> CONFIDENTIALITY NOTICE AND DISCLAIMER : This telecommunication, including any and all attachments, contains confidential information intended only for the person(s) to whom it is addressed. Any dissemination, distribution, copying or disclosure is strictly prohibited and is not a waiver of confidentiality. If you have received this telecommunication in error, please notify the sender immediately by return electronic mail and delete the message from your inbox and deleted items folders. This telecommunication does not constitute an express or implied agreement to conduct transactions by electronic means, nor does it constitute a contract offer, a contract amendment or an acceptance of a contract offer. Contract terms contained in this telecommunication are subject to legal review and the completion of formal documentation and are not binding until same is confirmed in writing and has been signed by an authorized signatory.

3 5

Pi in-place upgrade with weird side-effects
by Giles Orr 30 Nov '21

30 Nov '21

I have two Raspberry Pi 4s. Both had Raspbian 10 (buster) installed on them initially, and both have now had in-place upgrades to Raspberry Pi OS 11 (bullseye). I'm aware that the original and new OSes aren't quite the same beast (Raspbian vs. Raspberry Pi OS). For the most part, this went fine, but has left some unfortunate artifacts. The first (that I know of, there may be many others) is /etc/apt/sources.list.d/raspi.list, which includes one active line: deb http://archive.raspberrypi.org/debian/ buster main Notice that this is pointing to buster. Which made me do some digging: # for package in $(dpkg-query -W --showformat='${Package}\n'); do if apt info $package 2> /dev/null | grep -q buster ; then echo $package ; fi ; done chromium-browser chromium-browser-l10n chromium-codecs-ffmpeg-extra libraspberrypi-bin libraspberrypi-dev libraspberrypi-doc libraspberrypi0 raspberrypi-bootloader raspberrypi-kernel rng-tools-debian These are all packages that are installed from the buster repository, and they seem like some seriously important packages (the kernel? the bootloader?) that are being held back at an older version of the OS. I would rather not re-install the OS, so any suggestions about what's going on or how to fix it (or if that's necessary) would be appreciated. -- Giles https://www.gilesorr.com/ gilesorr(a)gmail.com

1 0

war story: Fedora 35 and Nvidia GTX 650
by D. Hugh Redelmeier 27 Nov '21

27 Nov '21

I've been using a GTX 650 on my desktop for over seven years. This may surprise you since I am an open-source fan. The reason I chose it is that it was all I had that would drive my UltraHD TV at full resolution. Things are better now (not as much as one would like) but I'm still using the same setup. The open source Nouveau driver isn't up to the job. It wouldn't drive the monitor seven years ago. I try it every few years but it never does the job I need. Currently it seems to be OK until I start my million-tab Firefox session. I upgraded from Fedora 34 to 35, using the in-place dnf method. <https://docs.fedoraproject.org/en-US/quick-docs/dnf-system-upgrade/> After that, the nvidia driver would not run. I thrashed about for a considerable amount of time and I cannot explain what happened, but I got it working. Here are some of the things I found useful. I used the RPM Fusion repository for the drivers. (Red Hat doesn't distribute closed source drivers.) This document is useful: <https://rpmfusion.org/Howto/NVIDIA#Legacy_GeForce_600.2F700> When nvidia drivers fail to load, the log is not at all helpful. (One reason can be that the firmware has Secure Boot enabled, but that was not my problem.) Nvidia considers my card "legacy" so the mainstream drivers no longer support it. The RPM Fusion drivers for Fedora 34 and earlier did but not for Fedora 35. Basically, I have to switch to the nvidia-470xx series of drivers. Hint: I think you only want one generation of nvidia driver installed on a machine (I could be wrong). If you are changing generations, rip the former ones out before installing the new ones. In the Fedora 34 time frame, something else when wrong (see an earlier war story) so I created a new F34 installation that shared the /home partition with my old one. I had two root partitions, one for the old F34 and another for the fresh F34. This was handy when I was experimenting with this problem. But it exposed my ignorance of grub2 in Fedora. I always thought that you could fix the grub configuration by sudo grub2-mkconfig -o /boot/efi/EFI/fedora/grub.cfg Not so since Fedora 34. Read this: <https://fedoraproject.org/wiki/GRUB_2> (It even covers the case where you have screwed it up like I did.)

1 0

Upgrading SSD -> NVMe
by William Park 27 Nov '21

27 Nov '21

Has anyone upgraded from SSD SATA to NVMe M.2? Spec says it's 10x the speed. What did you really see? --William

2 1

Swapping ThinkPad drive
by James Knott 27 Nov '21

27 Nov '21

I've bought a 1 TB SSD drive to replace the 750 GB drive in my ThinkPad. I was considering using the opportunity to change to UEFI. I'll be mounting the old drive in an external case so that I can copy over the partitions, one of them NTFS. Anything I should worry about in converting to UEFI? Any suggestions? tnx jk

6 7

[GTALUG-Announce] New Board and Officers
by Evan Leibovitch 26 Nov '21

26 Nov '21

Hello everyone, As you know, GTALUG held its Annual General Meeting earlier this month, and after that and a subsequent Board meeting I am happy to announce the coming year's Directors, Officers and supporting volunteers. Directors: - David Collyer-Brown - Alan Heighway - Evan Leibovitch - Warren McPherson - Alex Volkov Officers: - President: Evan - Treasurer: Warren - Secretary: Alan Operations Team: - High Redelmeier - Scott Sullivan I want to thank everyone for service to date; I am honored by the selection and hope to serve my role well. Please feel free to email me with any questions, speaker ideas or anything else on how we can make GTALUG better. As an initial move, GTALUG has created a Discord server where we hope to meet with all our friends who are using that platform. Thanks again, Evan Leibovitch, Toronto Canada @evanleibovitch / @el56 --- GTALUG Announce mailing list announce(a)gtalug.org https://gtalug.org/mailman/listinfo/announce

5 7

Swapping SSD Drives
by Jim Ruxton 24 Nov '21

24 Nov '21

I am currently booting off an SSD drive that boots both Linux (Ubuntu 20.04) and Windows. The partition scheme is shown here: https://drive.google.com/file/d/1zED6cbDncGz-fd1KF2bMjYEf5v3VWSOn/view?usp=… <https://drive.google.com/file/d/1zED6cbDncGz-fd1KF2bMjYEf5v3VWSOn/view?usp=…> I want to get a larger SSD and replace it. Can I just clone this drive to a larger one using Clonezilla? My thought is to put Clonezilla on a USB drive and then clone the current NVMe to another USB key. Then remove current SSD and replace with a larger one. Boot back into Clonezilla and clone what is on the second USB key to the new SSD in the laptop. Then reboot and voila. Is it that straightforward? I guess another option is to get a NVMe to USB adapter and clone directly to that? Is that a better solution? I've never used Clonezilla before so in unfamiliar waters here. I also have a large HDD in the machine so maybe could copy the current SDD to that as well? Feeling a little lost as to best way to move forward. Thanks for any tips. Jim

3 8