Date: Sat, 21 Nov 2020 00:02:42 +0100 From: Mateusz Guzik <mjguzik@gmail.com> To: mike tancsa <mike@sentex.net> Cc: Philip Paeps <philip@freebsd.org>, "Bjoern A. Zeeb" <bz@freebsd.org>, netperf-admin@freebsd.org, netperf-users@freebsd.org, Allan Jude <allanjude@freebsd.org> Subject: Re: zoo reboot Friday Nov 20 14:00 UTC Message-ID: <CAGudoHE%2BxjHdBQAD3cAL84=k-kHDsZNECBGNNOn2LsStL5A7Dg@mail.gmail.com> In-Reply-To: <a1ef98c6-e734-1760-f0cb-a8d31c6acc18@sentex.net> References: <1f8e49ff-e3da-8d24-57f1-11f17389aa84@sentex.net> <5A5094BC-D417-4BA6-97E2-7CB522B51368@FreeBSD.org> <4ec6ed6f-b3b4-22ae-e1ec-93a46f3d88ea@sentex.net> <d2ffd0f1-1dd8-dc6b-9975-93f20d7974a4@sentex.net> <dc8fed75-0262-c614-3292-6b8ce5addcfc@sentex.net> <0ddec867-32b5-f667-d617-0ddc71726d09@sentex.net> <CAGudoHHNN8ZcgdkRSy0cSaPA6J9ZHVf%2BBQFiBcThrtQ0AMP%2BOw@mail.gmail.com> <5549CA9F-BCF4-4043-BA2F-A2C41D13D955@freebsd.org> <ad81b5f3-f6de-b908-c00f-fb8d6ac2a0b8@sentex.net> <CAGudoHETJZ0f_YjmCcUjb-Wcf1tKhSF719kXxXUB3p4RB0uuRQ@mail.gmail.com> <CAGudoHH=H4Xok5HG3Hbw7S=6ggdsi%2BN4zHirW50cmLGsLnhd4g@mail.gmail.com> <270b65c0-8085-fe2f-cf4f-7a2e4c17a2e8@sentex.net> <CAGudoHFLy2dxBMGd2AJZ6q6zBsU%2Bn8uLXLSiFZ1QGi_qibySVg@mail.gmail.com> <a716e874-d736-d8d5-9c45-c481f6b3dee7@sentex.net> <CAGudoHELFz7KyzQmRN8pCbgLQXPgCdHyDAQ4pzFLF%2BYswcP87A@mail.gmail.com> <163d1815-fc4a-7987-30c5-0a21e8383c93@sentex.net> <CAGudoHF3c1e2DFSAtyjMpcrbfzmMV5x6kOA_5BT5jyoDyKEHsA@mail.gmail.com> <a1ef98c6-e734-1760-f0cb-a8d31c6acc18@sentex.net>
next in thread | previous in thread | raw e-mail | index | archive | help
should you go trhough with it please make sure zsh is installed. then i can get in and install some other stuff as needed. On 11/20/20, mike tancsa <mike@sentex.net> wrote: > That looks good to my newbie eyes. I think tomorrow I will just bite > the bullet and put in a pair of 240G SSDs to boot from. I will install > HEAD onto them on another machine, and then put them in zoo, boot from > there and adjust the home directory mounts accordingly if thats OK with > everyone ? > > ---Mike > > On 11/20/2020 5:53 PM, Mateusz Guzik wrote: >> I ran this one-liner: >> >> gpart bootcode -b /boot/pmbr -p /boot/gptzfsboot -i 1 ada0 >> >> which according to https://wiki.freebsd.org/RootOnZFS/GPTZFSBoot >> should be fine. Hopefully Allan will know better. >> >> On 11/20/20, mike tancsa <mike@sentex.net> wrote: >>> Unfortunately no luck :( >>> >>> >>> >>> ZFS: i/o error - all block copies >>> unavailable >>> ZFS: can't read MOS of pool >>> zroot >>> gptzfsboot: failed to mount default pool >>> zroot >>> >>> >>> FreeBSD/x86 >>> boot >>> >>> >>> Whats odd is that it doesnt post all the drives.... >>> >>> Zoo predated EFI, so it was booting legacy BIOS. Are the boot blocks >>> that you installed assuming that ? >>> >>> On 11/20/2020 1:27 PM, Mateusz Guzik wrote: >>>> swap and boot partitions resized, the ada0p3 partition got removed >>>> from the pool and inserted back, it is rebuilding now: >>>> >>>> root@zoo2:~ # zpool status >>>> pool: zroot >>>> state: DEGRADED >>>> status: One or more devices is currently being resilvered. The pool >>>> will >>>> continue to function, possibly in a degraded state. >>>> action: Wait for the resilver to complete. >>>> scan: resilver in progress since Fri Nov 20 23:13:28 2020 >>>> 459G scanned at 1.00G/s, 291G issued at 650M/s, 3.47T total >>>> 0B resilvered, 8.17% done, 01:25:48 to go >>>> config: >>>> >>>> NAME STATE READ WRITE >>>> CKSUM >>>> zroot DEGRADED 0 0 >>>> 0 >>>> mirror-0 DEGRADED 0 0 >>>> 0 >>>> replacing-0 DEGRADED 0 0 >>>> 0 >>>> 1517819109053923011 OFFLINE 0 0 >>>> 0 was /dev/ada0p3/old >>>> ada0p3 ONLINE 0 0 >>>> 0 >>>> ada1 ONLINE 0 0 >>>> 0 >>>> mirror-1 ONLINE 0 0 >>>> 0 >>>> ada3p3 ONLINE 0 0 >>>> 0 >>>> ada4p3 ONLINE 0 0 >>>> 0 >>>> mirror-2 ONLINE 0 0 >>>> 0 >>>> ada5p3 ONLINE 0 0 >>>> 0 >>>> ada6p3 ONLINE 0 0 >>>> 0 >>>> special >>>> mirror-3 ONLINE 0 0 >>>> 0 >>>> gptid/db15e826-1a9c-11eb-8d25-0cc47a1f2fa0 ONLINE 0 0 >>>> 0 >>>> mfid1p2 ONLINE 0 0 >>>> 0 >>>> >>>> errors: No known data errors >>>> >>>> One pickle: i did 'zpool export zroot' to replace the drive, otherwise >>>> zfs protested. subsequent zpool import was done slightly carelessly >>>> and it mounted over /, meaning i lost access to original ufs. Should >>>> there be a need to boot from it again someone will have to boot single >>>> user and make sure to comment out swap in /etc/fstab or we will have >>>> to replace the drive again. >>>> >>>> That said, as I understand we are in position to take out the ufs >>>> drive and reboot to be back in business. >>>> >>>> The ufs drive will have to be mounted somewhere to sort out that swap. >>>> >>>> On 11/20/20, mike tancsa <mike@sentex.net> wrote: >>>>> On 11/20/2020 1:00 PM, Mateusz Guzik wrote: >>>>>> So this happened after boot: >>>>>> >>>>>> root@zoo2:/home/mjg # swapinfo >>>>>> Device 1K-blocks Used Avail Capacity >>>>>> /dev/ada0p3 2928730500 0 2928730500 0% >>>>>> >>>>>> which i presume might have corrupted some of it. >>>>> Oh, that makes sense now. When it was installed in the back, the drive >>>>> posted as ada0. When we put it in zoo, it was on a farther down port, >>>>> hence it came up as ada7. I had to manually mount / off ada7p2. I >>>>> updated fstab so as not to do that again already. That mystery >>>>> solved. >>>>> >>>>> ---Mike >>>>> >>>>> >>>>>> Allan pasted some one-liners to resize the boot and swap partition. >>>>>> >>>>>> With your permission I would like to run them and then offline/online >>>>>> the disk to have it rebuild. >>>>>> >>>>>> As for longer plans what to do with it i think that's a different >>>>>> subject, whatever new drives end up being used I'm sure the FreeBSD >>>>>> Foundation can reimburse you with no difficulty. >>>>>> >>>>>> >>>>>> On 11/20/20, mike tancsa <mike@sentex.net> wrote: >>>>>>> Its a bit of an evolutionary mess the current state of zoo. I >>>>>>> wonder >>>>>>> if >>>>>>> we are better off re-installing the base OS fresh on a pair of SSD >>>>>>> drives and have the base OS on it and leave all the user data on the >>>>>>> current "zroot"... Considering 240G SSDs are $35 CDN it might be >>>>>>> easier >>>>>>> to just install fresh on it and not have to worry about resizing >>>>>>> etc. >>>>>>> >>>>>>> ---Mike >>>>>>> >>>>>>> On 11/20/2020 12:49 PM, Mateusz Guzik wrote: >>>>>>>> On 11/20/20, Mateusz Guzik <mjguzik@gmail.com> wrote: >>>>>>>>> CC'ing Allan Jude >>>>>>>>> >>>>>>>>> So: >>>>>>>>> >>>>>>>>> pool: zroot >>>>>>>>> state: DEGRADED >>>>>>>>> status: One or more devices could not be opened. Sufficient >>>>>>>>> replicas >>>>>>>>> exist >>>>>>>>> for >>>>>>>>> the pool to continue functioning in a degraded state. >>>>>>>>> action: Attach the missing device and online it using 'zpool >>>>>>>>> online'. >>>>>>>>> see: https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-2Q >>>>>>>>> scan: scrub repaired 0B in 05:17:02 with 0 errors on Tue Aug 18 >>>>>>>>> 15:19:00 >>>>>>>>> 2020 >>>>>>>>> config: >>>>>>>>> >>>>>>>>> NAME STATE READ >>>>>>>>> WRITE >>>>>>>>> CKSUM >>>>>>>>> zroot DEGRADED 0 >>>>>>>>> 0 >>>>>>>>> 0 >>>>>>>>> mirror-0 DEGRADED 0 >>>>>>>>> 0 >>>>>>>>> 0 >>>>>>>>> 1517819109053923011 UNAVAIL 0 >>>>>>>>> 0 >>>>>>>>> 0 was /dev/ada0p3 >>>>>>>>> ada1 ONLINE 0 >>>>>>>>> 0 >>>>>>>>> 0 >>>>>>>>> mirror-1 ONLINE 0 >>>>>>>>> 0 >>>>>>>>> 0 >>>>>>>>> ada3p3 ONLINE 0 >>>>>>>>> 0 >>>>>>>>> 0 >>>>>>>>> ada4p3 ONLINE 0 >>>>>>>>> 0 >>>>>>>>> 0 >>>>>>>>> mirror-2 ONLINE 0 >>>>>>>>> 0 >>>>>>>>> 0 >>>>>>>>> ada5p3 ONLINE 0 >>>>>>>>> 0 >>>>>>>>> 0 >>>>>>>>> ada6p3 ONLINE 0 >>>>>>>>> 0 >>>>>>>>> 0 >>>>>>>>> special >>>>>>>>> mirror-3 ONLINE 0 >>>>>>>>> 0 >>>>>>>>> 0 >>>>>>>>> gptid/db15e826-1a9c-11eb-8d25-0cc47a1f2fa0 ONLINE 0 >>>>>>>>> 0 >>>>>>>>> 0 >>>>>>>>> mfid1p2 ONLINE 0 >>>>>>>>> 0 >>>>>>>>> 0 >>>>>>>>> >>>>>>>>> errors: No known data errors >>>>>>>>> >>>>>>>>> # dmesg | grep ada0 >>>>>>>>> Trying to mount root from ufs:/dev/ada0p2 [rw]... >>>>>>>>> ada0 at ahcich0 bus 0 scbus0 target 0 lun 0 >>>>>>>>> ada0: <WDC WD3003FZEX-00Z4SA0 01.01A01> ACS-2 ATA SATA 3.x device >>>>>>>>> ada0: Serial Number WD-WCC137TALF5K >>>>>>>>> ada0: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) >>>>>>>>> ada0: Command Queueing enabled >>>>>>>>> ada0: 2861588MB (5860533168 512 byte sectors) >>>>>>>>> ada0: quirks=0x1<4K> >>>>>>>>> Mounting from ufs:/dev/ada0p2 failed with error 2; retrying for 3 >>>>>>>>> more >>>>>>>>> seconds >>>>>>>>> Mounting from ufs:/dev/ada0p2 failed with error 2. >>>>>>>>> vfs.root.mountfrom=ufs:/dev/ada0p2 >>>>>>>>> GEOM_PART: Partition 'ada0p3' not suitable for kernel dumps (wrong >>>>>>>>> type?) >>>>>>>>> ZFS WARNING: Unable to attach to ada0p3. >>>>>>>>> ZFS WARNING: Unable to attach to ada0p3. >>>>>>>>> ZFS WARNING: Unable to attach to ada0p3. >>>>>>>>> ZFS WARNING: Unable to attach to ada0p3. >>>>>>>>> ZFS WARNING: Unable to attach to ada0p3. >>>>>>>>> ZFS WARNING: Unable to attach to ada0p3. >>>>>>>>> >>>>>>>>> # gpart show ada0 >>>>>>>>> => 34 5860533101 ada0 GPT (2.7T) >>>>>>>>> 34 6 - free - (3.0K) >>>>>>>>> 40 88 1 freebsd-boot (44K) >>>>>>>>> 128 3072000 2 freebsd-swap (1.5G) >>>>>>>>> 3072128 5857461000 3 freebsd-zfs (2.7T) >>>>>>>>> 5860533128 7 - free - (3.5K) >>>>>>>>> >>>>>>>>> Running naive dd if=/dev/ada0p3 works, so I don't know what zfs >>>>>>>>> complains >>>>>>>>> about. >>>>>>>>> >>>>>>>> Also note Philip's point boot partition of 44k. Is that too small >>>>>>>> now? >>>>>>>> >>>>>>>>> On 11/20/20, mike tancsa <mike@sentex.net> wrote: >>>>>>>>>> On 11/20/2020 11:40 AM, Philip Paeps wrote: >>>>>>>>>>> On 2020-11-21 00:04:19 (+0800), Mateusz Guzik wrote: >>>>>>>>>>> >>>>>>>>>>>> Oh, that's a bummer. I wonder if there is a regression in the >>>>>>>>>>>> boot >>>>>>>>>>>> loader though. >>>>>>>>>>>> >>>>>>>>>>>> Does the pool mount if you boot the system from a cd/over the >>>>>>>>>>>> network/whatever? >>>>>>>>>>> It's worth checking if the freebsd-boot partition is large >>>>>>>>>>> enough. >>>>>>>>>>> I >>>>>>>>>>> noticed during the cluster refresh that we often use 108k for >>>>>>>>>>> freebsd-boot but recent head wants 117k. I've been bumping the >>>>>>>>>>> bootblocks to 236k. >>>>>>>>>>> >>>>>>>>>>> So far, all the cluster machines I've upgraded booted though .. >>>>>>>>>>> so >>>>>>>>>>> ... >>>>>>>>>>> I might be talking ex recto. :) >>>>>>>>>>> >>>>>>>>>> I put in an ssd drive and booted from it. One of the drives might >>>>>>>>>> have >>>>>>>>>> gotten loose or died in the power cycles, but there is still >>>>>>>>>> redundancy >>>>>>>>>> and I was able to mount the pool. Not sure why it cant find the >>>>>>>>>> file >>>>>>>>>> ? >>>>>>>>>> >>>>>>>>>> root@zoo2:~ # diff /boot/lua/loader.lua /mnt/boot/lua/loader.lua >>>>>>>>>> 29c29 >>>>>>>>>> < -- $FreeBSD$ >>>>>>>>>> --- >>>>>>>>>>> -- $FreeBSD: head/stand/lua/loader.lua 359371 2020-03-27 >>>>>>>>>>> 17:37:31Z >>>>>>>>>> freqlabs $ >>>>>>>>>> root@zoo2:~ # >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> % ls -l /mnt/boot/lua/ >>>>>>>>>> total 110 >>>>>>>>>> -r--r--r-- 1 root wheel 4300 Nov 20 08:41 cli.lua >>>>>>>>>> -r--r--r-- 1 root wheel 3288 Nov 20 08:41 color.lua >>>>>>>>>> -r--r--r-- 1 root wheel 18538 Nov 20 08:41 config.lua >>>>>>>>>> -r--r--r-- 1 root wheel 12610 Nov 20 08:41 core.lua >>>>>>>>>> -r--r--r-- 1 root wheel 11707 Nov 20 08:41 drawer.lua >>>>>>>>>> -r--r--r-- 1 root wheel 2456 Nov 20 08:41 gfx-beastie.lua >>>>>>>>>> -r--r--r-- 1 root wheel 2235 Nov 20 08:41 gfx-beastiebw.lua >>>>>>>>>> -r--r--r-- 1 root wheel 1958 Nov 20 08:41 gfx-fbsdbw.lua >>>>>>>>>> -r--r--r-- 1 root wheel 2413 Nov 20 08:41 gfx-orb.lua >>>>>>>>>> -r--r--r-- 1 root wheel 2140 Nov 20 08:41 gfx-orbbw.lua >>>>>>>>>> -r--r--r-- 1 root wheel 3324 Nov 20 08:41 hook.lua >>>>>>>>>> -r--r--r-- 1 root wheel 2395 Nov 20 08:41 loader.lua >>>>>>>>>> -r--r--r-- 1 root wheel 2429 Sep 24 09:09 logo-beastie.lua >>>>>>>>>> -r--r--r-- 1 root wheel 2203 Sep 24 09:09 logo-beastiebw.lua >>>>>>>>>> -r--r--r-- 1 root wheel 1958 Sep 24 09:09 logo-fbsdbw.lua >>>>>>>>>> -r--r--r-- 1 root wheel 2397 Sep 24 09:09 logo-orb.lua >>>>>>>>>> -r--r--r-- 1 root wheel 2119 Sep 24 09:09 logo-orbbw.lua >>>>>>>>>> -r--r--r-- 1 root wheel 14201 Nov 20 08:41 menu.lua >>>>>>>>>> -r--r--r-- 1 root wheel 4299 Nov 20 08:41 password.lua >>>>>>>>>> -r--r--r-- 1 root wheel 2227 Nov 20 08:41 screen.lua >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> >>>>>>>>> -- >>>>>>>>> Mateusz Guzik <mjguzik gmail.com> >>>>>>>>> >>> >> > -- Mateusz Guzik <mjguzik gmail.com>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAGudoHE%2BxjHdBQAD3cAL84=k-kHDsZNECBGNNOn2LsStL5A7Dg>