00:07:33 woop 00:07:41 RELEASE released 00:11:48 Hm, schedule said it'd not be until December 2nd. 00:40:09 It's not released until the announcement goes out to freebsd-announce@. 00:44:58 upgrade time! 00:57:25 :o 01:10:58 Welp, hope that the "sneak peek" will be the actual RELEASE release... 01:11:23 I'm most curious about pkgbase, because all the docs are still referring to it as experimental. 01:12:51 It sadly still has a fair amount of problems. I'm not jumping ship to that anytime soon. 01:13:39 pkgbase being complete might have compelled me to jump sooner, but I feel like I'll be staying on 14.3 for a while yet. 01:25:05 is that something you have to use? i thought freebsd-update is still part of 15 01:25:47 V_PauAmma_V: it's available for download. i mean the builds started today according to the release schedule and the RCs were out (around) the same as the dates listed 01:26:33 they probably just don't add it to the main download page until after announcement? 01:32:28 The period between now and until the announcement means there is on-going testing from Colin Percival for that RELEASE build. In those (perhaps rare) cases where it turns out to be problematic, a newer build may show up, leaving you in the dust with an unsupported build. 01:33:01 So you take a risk in trying it out. 01:33:07 Patience is a virtue! 01:33:13 Yep 02:04:29 Ah I'll wait then 02:22:47 how to fix this? 02:22:47 root@Install:~ # ls /dev/ada1s1* 02:22:47 /dev/ada1s1 /dev/ada1s1a /dev/ada1s1b 02:22:47 root@Install:~ # gpart show ada1s1 02:22:48 => 0 20971520 ada1s1 BSD (10G) 02:22:50 0 4194304 1 freebsd-ufs (2.0G) 02:22:53 4194304 16777216 2 freebsd-zfs (8.0G) 02:22:55 root@Install:~ # mount ada1s1a /mnt/ 02:22:56 mount: ada1s1a: No such file or directory 02:34:36 Chip1972: Give that ada1s1a a full path. 03:05:20 how to write bootloader to mbr? 03:24:57 how to fix this? 03:24:57 root@Install:~ # gpart show ada1s1 03:24:57 => 0 20971520 ada1s1 BSD (10G) 03:24:57 0 4194304 1 freebsd-ufs (2.0G) 03:24:59 4194304 16777216 2 freebsd-zfs (8.0G) 03:25:00 root@Install:~ # gpart set -a active -i 1 ada1s1 03:25:02 gpart: attrib 'active': Device not configured 03:33:56 Chip1972, Seeing that device ada1s1 it looks very strange to me. That's already a slice of ada1. Is that really what you intend? How was that created? What does "gpart show ada1" and "gpart show -l ada1" say? 03:34:58 And with that I don't think it is possible to set the active flag on a slice of a slice of a device. 03:42:55 Full disclosure I am not an expert on this topic. 03:46:37 rwp: I want to boot my manual installation. The boot is stalled "booting from hard disk" 03:48:28 Chip1972, I'll point you to this reference as a good place to start: https://wiki.freebsd.org/MasonLoringBliss/ZFSandGELIbyHAND 04:05:15 rwp: I am using MBR 04:06:34 But there is no freebsd-boot slice listed in your output. 04:08:32 Chip1972, This is another good reference: https://klarasystems.com/articles/the-freebsd-boot-process/ 04:09:41 Unless you absolutely must do it otherwise I strongly suggest using BIOS+GPT+ZFS. 04:25:15 I have a 1TB HD with MBR data. 04:38:24 The Klara Systems article I reference includes examples of installing boot code for the MBR case. 04:39:18 But I am still skeptical that one can nest partition tables like that and boot a nested slice that resides in a device slice. Maybe it is possible. But if so it is still highly atypical. 04:41:31 Normally with MBR on an ada0 device one might "gpart bootcode -b /boot/boot0 ada0" to install the boot code there and "gpart set -a active -i 1 ada0" to set it as the active partition. But your message showed "gpart set -a active -i 1 ada1s1" trying to set it upon a partition slice of a slice of the device. I don't think that can work. 04:52:01 hernan604: 04:52:06 FreeBSD drink.unix.beer 15.0-RELEASE FreeBSD 15.0-RELEASE releng/15.0-7aedc8de6446 GENERIC amd64 04:54:27 rwp: yes... I tested a default unstallation. and ada0s1 is the boot partition 04:56:05 If ada0s1 is the boot partition then you don't set -i 1 ada0s1. That should be "gpart set -a active -i 1 ada1" because -i 1 already selects the first slice of it. 04:57:05 The gpart -i option is the index of the target partition so -i 1 maps to s1. 06:06:32 i had to use a usb ethernet card to update to 15, not realzing that dhclient ue0 updates resolv.conf making bastille not work 08:48:47 is 15R looking good? 09:42:10 what is in the "optional" set on a 15.0 install? 11:58:54 Hi! Does anyone know if there is a port of sqlite_rsync? https://sqlite.org/rsync.html 13:01:39 is there a way to upgrade hard drive firmwares in fbsd? 14:49:06 silly question perhaps but is there documentation yet about using pkgbase for jails? 14:53:37 Macer: Yes. with nvmectl 14:54:22 or nvmecontrol rather 16:25:07 I meant platters. 16:25:33 It doesn’t matter. My disks don’t have an available firmware upgrade anyways. 16:49:09 i do not update nvme firmware on already working system. neither i do upgrade zfs pools. 2 goldens rules 17:24:08 i had to upgrade my pool. i want to use raidz expansion :) 17:24:25 although i do wish you could disable the upgrade message 17:24:57 or at least make it smaller.. like poolname (Upgrade) instead of the paragraph it gives 17:25:18 or maybe just a flag to show if you can do it 18:49:48 I have gotten so used to seeing the zpool status upgrade message that I mentally ignore it now. I am never in a hurry to upgrade the pools because it prevents being able to use Boot Environments to recovery boot the previous OS version. And the previous zfs features are sufficient for my use. 19:03:01 omg. i thought this barracuda wasn't SMR.. but i guess it is.. this resilver is going to take a week 19:04:02 i'll be better off just ordering a good disk.. waiting... and replacing with that 19:06:43 Friends don't let friends use SMR drives in a RAID. 19:08:54 Macer, I asked but didn't hear if this storage array you are working on is a test setup testing expansion or if this is an in use array that you are trying to expand. You said you had a previous disk failure? 19:11:01 no. i was trying to add a disk and one of the old drives broke while doing it lol 19:11:20 so the only one i had was a barracuda in a drawer but i didn't realize that an 8TB barracuda was smr so now it's stuck in the endless resilver 19:12:28 Might be best just to stop it, and remove it then, and wait for the replacement. 19:12:56 I am simply curious... What is the structural layout of the array? It's a mirror and operating on only one drive now? 19:13:04 yeah probably 19:13:08 it's raidz2 19:13:19 so i have another disk that can fail before i have to start getting nervous 19:14:15 It's raidz2 so at least 3 disks in it at one time, now with one failed and removed. 2 drives still working. State will be degraded. Gotcha. 19:14:26 SMR is such garbage.. i'm trying to think what kind of use case they'd be used. i guess maybe for capturing low bandwidth video footage from cameras where zero performance is necessary? 19:14:37 yeah. it's degraded now. 19:15:20 I would walk carefully around it so that neither of the remaining drives get bumped! 19:16:25 Most consumers buying stuff like this have no idea what they are buying. They don't know anything about CMR or SMR or anything. They just see USD$$ and a 8TB or some combination and go, that's the cheapest one, I am buying it. 19:17:35 As to what it would actually be good for, well, I guess it would be okay as a single drive in a USB that is sometimes connected and then an image dump written to it. That would start at the beginning of the drive and proceed and that is relatively efficient to do on SMR drives. It's the random access which is terrible. 19:18:45 So actually video storage could be okay with it. Maybe. I am not using one but maybe in a video storage system which is just writing huge files linearly then it might be okay on an SMR. 19:19:43 yeah at the time i didn't know that SMR was hot garbage since it was relatively new and the drives were so cheap 19:20:00 i also wasn't sure if the 8TB barracudas were SMR at all. i just figured they were cheap consumer grade disks 19:20:05 now they're a problem :/ 19:31:58 Macer: rough, I once made the same mistake. Newegg did not have the drives labeled as SMR x_x 19:33:07 eventually one failed and I couldn't resilver the ZFS pool correctly, since drives started to throw all sorts of controller errors with higher random write load, ended having to cry and buy new drives for a totally new pool. Fortunately I was able to mount the old pool in a degraded state and save data 19:44:34 General Hint: If a drive fails in a zfs array don't immediately remove it and then replace it. Instead add in the new replacement drive. Use the zfs replace command to silver the new drive while the old one is still there and partially working. It's a "make before break" process and somewhat safer. 19:45:39 rwp: does it still make use of a failed drive somehow? 19:45:40 Then after zfs has replaced the drive the failed drive is no longer in use then remove it from the system. Unless the drive is completely dead then it still holds some of the redundancy and can still provide data blocks for the working part of the drive if it has not completely failed. 19:45:53 even if reads are failing? 19:45:57 hodapp, Yes. 19:46:42 If the drive has completely failed and nothing is working then might as well remove it. But mostly when spinning drives fail there is still partial ability to read sectors that are still working and zfs will read those if needed. 19:47:10 With SSDs though for me a more typical failure is that the SSD controller fails and then the entire device is completely dead. 19:49:16 At one time I had a drive that was giving read errors on some data blocks and I used GNU ddrescue to clone the failing drive onto a good drive. I did that in a different system. Let it run for more than 24 hours before it had scraped everything it could off the failing drive. Then I put the good drive with the copy of that data back into the zfs array. Then ran a scrub. That re-silvered all of the missing data blocks back onto the 19:49:25 new good drive. 19:49:56 my most bizarre failure was on a cheap SSD that would just fall off the bus after a few weeks 19:50:29 Would it come back after a power off power on cycle? And then repeat again after a few weeks? 19:52:19 yes, but for whatever reason this manifested as: the device getting the wrong address because DHCP was never released properly 19:52:38 That is a bizarre stream of events. 19:59:50 it took me a long time to connect "box is on the wrong IP" with "drive fell off the bus" 20:01:35 I think it was finally by chance I caught it happening after I plugged in a monitor and was tailing logs or something 20:53:50 rtyler: yeah. it makes me wonder if raidz expansion is even possible with this pool because of that and i have it hanging out in the background waiting on the resilver 20:53:58 and one drive just removed itself from the pool 20:54:48 which is odd because it was actually moving along quite well since i guess the writes are awkward during reflow 20:56:32 expand: expansion of raidz2-0 in progress since Thu Nov 27 22:35:46 2025 20:56:34 12.2T / 69.2T copied at 88.1M/s, 17.63% done, paused for resilver or clear 20:57:10 it hit 12 TB and then a drive just up and completely died. like died died because pulling it and putting it in another bay sure didn't help matters. freebsd wouldn't even boot until i yanked it out with nonstop controller errors 20:58:15 829G resilvered, 15.02% done, 5 days 01:24:32 to go 20:58:28 i suppose i can just grind through it for 5 days since amazon can't get me a drive here until wednesday for some reason anyways 21:02:54 hodapp: what on earth does dhcp have to do with an ssd falling off the bus? lol 21:05:34 If it were me then before I expanded an array I would have it scrubbed and re-silvered so that the array was online and not degraded. Then I would expand it. 21:10:34 i did scrub it. just recently. this seems like like catastrophic drive failure 21:10:53 it happened during the expansion 21:11:30 it is what it is. i'm down 1 disk of parity that's rebuilding so i'll just wait it out and hopefully it completes without issue.. but i think i'll definitely work on making sure i remove all the SMR drives little by little. 21:12:01 which i guess aren't bad when you're simply using them. but expanding them and restoring a pool with them is not happening. 21:32:11 after updating my desktop I noticed that install the gpu firmware kmods manually it doesnt install beige or renoir 21:32:39 do I had to remove all gpu stuff, and just let fwget install what it needs 21:32:59 after about an hr kldload amdgpu finally didint panic my system