03:36:43 rrsync is no longer part of the rsync backage all my daily backups are broke reeEEEEeeee 03:48:52 ok i installed rsync-python but it conflicted with and deleted mariadb lol 03:50:22 this is lulzy https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=286073 07:31:09 I thought I could limit the # of cpu cores by setting `hw.ncpu="2"` in loader.conf 07:31:41 or do I recall wrongly and that only works for memory? 07:33:30 that's a read only var 07:33:57 perhaps this can reach you to your goal, https://man.freebsd.org/cgi/man.cgi?cpuset 08:17:47 dch: yes, that should work 08:19:36 dch: although maybe this is only on ARM 08:19:50 i don't see anything in sys/amd64 that fetches the tunable 09:18:18 iRobbery: not in loader.conf its not 09:18:48 ivy: do you look for something like egrep GETSYSCTL.+hw.ncpu ? 09:19:10 dch: TUNABLE.*ncpu 09:19:13 * dch finds GETSYSCTL("kern.smp.maxcpus", maxcpu); 09:19:33 ivy: thanks, that a neat trick 09:20:24 I worked backwards from that to find the smp(4) manpage 09:23:10 use git grep btw, it's often a lot faster than grep(1) :-) 09:26:57 I use ripgrep, its 2x faster than git grep 09:36:57 what's the difference between cpuset and maxcpus ? I understand the first is non-permanent? 10:23:17 so yes hw.ncpus appears to be an ARM only thing, and kern.smp.maxcpus isn't writable as far as I can tell 10:23:32 would be nice to have this, so I logged a PR 11:51:14 dch: ah good catch, ill remember that from now 14:34:42 ek: hi! 14:35:27 just read ur mail, gonna ask the postmaster 14:37:01 very funny that my "test" email got delivered (altough after a few hours) 14:37:37 and i tried sending a mail to pi@ directly, but this one got delayed again for same reason -_- 14:51:07 riad, maybe test your E-mail config at learndmarc.com 14:52:50 CrtxReavr: thanks, will try 14:52:53 really cool site 14:53:31 Indeed. 14:54:02 My SPF & DKIM passed (Google Apps domain), but DMARC failed. . . 14:54:34 But I had to regenerate a different DKIM key, amongst other work, to get my DMARC working correctly. 14:54:47 It was actually a longer story than that. 14:54:55 Needed some help from #email. 14:57:23 Vaultwarden is now broken on 13.x and one of its deps is broken on 15.x and 16.x. I hope capnproto will work again soon. 15:42:37 eek. hopefully I don't need to update during this period where it must be fixed 15:51:14 CrtxReavr: that sounds like a bit of a pain in the ass... 15:51:31 dmarc is passing for me though 15:52:51 Is SPF & DKIM passing? 16:00:53 CrtxReavr: 16:00:55 yes 16:02:21 ek thinks ipv6 routing is failing somethere along the line 16:02:25 might be possible 16:03:07 riad: Do you not have access to the mail server? 16:03:34 ek: i do not 16:03:44 but email to admin is sent 16:04:31 I just find it strange that it doesn't also attempt via v4 if v6 fails. Might be a shoddy SMTP setup or something. 18:01:44 vkarlsen Oh wow. Thanks for letting me know. 18:44:17 hey 18:45:56 i have FreeBSD 15.0 with ZFS pool disk. One of my disk failed so i'm trying to replace it. I did that many times in the past without any problem but when I stop the computer, change the failed disk and boot it again I got "zio_read error: 5" and "ZFS: i/o error - all block copies unavailable", "ZFS: failed to read pool disk directory object" and it can't boot 18:46:16 when I remove the new disk and put the old one in place, the computer accepts to boot... 18:46:38 zpool status only complains about the faulty disk and not the others 18:49:09 My first thought is that the good disk was replaced instead of the faulty disk. 18:50:11 it's a RAID 5 with 5 disks 18:50:15 4 disks* 18:51:03 I also checked that the serial number of the faulty disk is the same as the one returned by smartctl 18:53:04 And unfortunately the hardware does not have hot swap bays requiring you to power down? :-( 18:55:15 Reading your report carefully the main problem is that it can't boot. Makes me think maybe the boot loader code is on the disk being removed maybe? I might boot the installer iso and then use it to examine things in the new configuration using the "zpool import" command. 18:57:35 titou: yeah, that'd be my guess, the failed drive is the first one, and it's the one with the loader 18:57:53 swap drive 0 and 1 and i would expect it to boot 18:59:48 I was thinking that too. 19:01:01 Booting the install ISO allows getting a live boot shell. And in the live boot shell one can examine the array with "zpool import" to see what arrays are available and what devices are in them and the state of the collection. 19:02:30 Also another possibility is that the currently failed drive is not fully failed but is providing just enough data blocks to make things work because the remaining drive is also faulting. I had that situation. I could not /remove/ the failing drive because that put the remaining disks below the redundancy limit and the array was FAULTED. 19:04:04 You might need to add in the new replacement drive along with the existing drives and then use the zpool replace to silver it while the quorum of drives is still in place. 19:04:14 Please post the output of zpool status? 19:07:03 hum 19:07:10 I installed the bootloader on the four disks 19:07:38 by using: gpart bootcode -b /boot/pmbr -p /boot/gptzfsboot -i 2 adaX with X from 0 to 3 19:07:52 I also put the EFI file system on the four disks 19:09:37 since I restarted zpool status returns nothing.. i had checksum errors on the faulty disks. Smartctl returns no error.. and no error into dmesg too 19:10:08 I bought my four disks on the same time and they are dying one after the other.. silly barracuda :( 19:13:31 got a faulty zpool satus from daily mails: https://pastebin.com/HPxxjstU 19:15:28 I assume the faulty disk is ada3 with the cksum error count? The problem with the adaX device names is that they might change between reboots. So be sure to triple check the serial number of the drive. I use "smartctl -i /dev/ada3" for it and "smartctl -a /dev/ada3" to look at all of the data from it. 19:22:21 indeed I will double check 19:22:29 i restarted a zpool scrub and check again 19:22:32 thank you for your help 19:30:43 Is there a web based provides like the pkg plugin? 19:31:05 ex: What remote package provide a given file/glob 19:36:34 there is freshports.org, unless you were thinking of something else 19:43:05 ah that's right. Thanks. 20:25:09 titou, I know you are already scrubbing now but since it was only showing a smallish number of cksum errors the disk might not be actually fully bad yet. And sometimes errors can be caused by cable problems and power connector problems. By doing what you have already done of removing and replacing the drive that might rub the connectors and things might work better now. 20:25:52 I know you are already scrubbing but running a "zpool clear" to reset the counts and clear errors followed by a "zpool scrub" so that you know any new error counts are new since then would be appropriate. 20:27:50 Also I would run smart selftests and see the result. "smartctl -l error /dev/ada3" and "smartctl -l selftest /dev/ada3". At least a "short" test but depending upon the drive a long test may be useful. My latest drives I can't run the long test because it takes longer than a day to complete and the drive is consumed during that time. So I don't. But the short test is good. 20:29:00 SMART has never predicted a drive failure for me but if a drive is actually failing then it has always been accompanied by smart failures logged. But if an error has been a power connector and sometimes SATA cabling then sometimes the drive will log DMA failures that are really data transfer errors and I can ignore them when used with ZFS with redundancy. 20:30:01 rwp++ 20:38:16 rwp: thank you for these information 20:38:56 I'm agree with you that previous failed disks had smart errors.. and that's strange that i don't have smart errors in this case 20:39:15 i did zpool clear two or three times and errors were corrected during each scrub 20:39:29 but maybe the SATA connector is failing.. i'll check with another one 20:45:08 Also double check the power connector. I had a bad 4-pin MOLEX power connector previously that caused me trouble. If the drive's own smart data is error free then I suspect cables. 20:46:07 Additionally if the drive is actually producing errors then the kernel should be logging something in /var/log/messages system log file. 21:30:13 rwp: nothing into the logs.. 21:30:22 i will double check. Thanks a lot