XCP-ng
    • Categories
    • Recent
    • Tags
    • Popular
    • Users
    • Groups
    • Register
    • Login

    XCP-NG 8.3 freezes and / or reboot on HGR-SAP3 by Ovh / Using ZFS on Raid 50 / With deduplication

    Scheduled Pinned Locked Moved Hardware
    20 Posts 2 Posters 1.9k Views 2 Watching
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • olivierlambertO Offline
      olivierlambert Vates 🪐 Co-Founder CEO
      last edited by

      Hi,

      You need to add more memory in your Dom0

      Y 1 Reply Last reply Reply Quote 0
      • Y Offline
        yann_m @olivierlambert
        last edited by

        @olivierlambert

        Thank's.

        Is it just a lack of memory for my Dom0 ?
        What memory size do you recommend ?

        This is my production environment, so I'll have to wait before I can do that.

        1 Reply Last reply Reply Quote 0
        • olivierlambertO Offline
          olivierlambert Vates 🪐 Co-Founder CEO
          last edited by

          Because ZFS eats a lot of RAM, go with 32GiB in the Dom0.

          Y 1 Reply Last reply Reply Quote 0
          • Y Offline
            yann_m @olivierlambert
            last edited by yann_m

            @olivierlambert "
            Tank's a lot

            I'll test it as soon as I can restart my host,
            I've read a lot of messages saying that putting too much memory in Dom0 is counterproductive.
            I'll keep you posted.

            1 Reply Last reply Reply Quote 0
            • olivierlambertO Offline
              olivierlambert Vates 🪐 Co-Founder CEO
              last edited by

              It is, if you don't use ZFS. ZFS is a memory hog.

              1 Reply Last reply Reply Quote 0
              • Y Offline
                yann_m
                last edited by

                Last night, I put 32GiB on the Dom0.

                For information, since I restarted, it's consuming approximately 17 GiB.

                I'll see if the increased load expected this weekend and the next won't cause it to crash.

                1 Reply Last reply Reply Quote 0
                • olivierlambertO Offline
                  olivierlambert Vates 🪐 Co-Founder CEO
                  last edited by

                  Great, keep us posted 🙂

                  1 Reply Last reply Reply Quote 0
                  • Y Offline
                    yann_m
                    last edited by

                    Unfortunately, I'm really pessimistic about my technological choice.

                    Maybe I should never have installed ZFS on it, or perhaps I should have never enabled deduplication ?

                    I didn't have time to post this before, but Wednesday morning, after some major backups, I lost all my VMs for 3-5 minutes (lost, but no crash) due to overactivity on tapdisk.

                    Luckily I found them again after.

                    So I haven't lost anything except my sleep.

                    Here's an example of one of my 14 logs (14 VMs) (I've anonymized the UUIDs).

                    Sep 24 04:02:03: [04:02:03.958] tapdisk[728741]: log start, level 0
Sep 24 05:27:00: [05:27:00.943] tapdisk_vbd_check_progress: vhd:/var/run/sr-mount/myZFSraid50/diskVM1.vhd: watchdog timeout: pending requests idle for 624 seconds
Sep 24 05:27:00: [05:27:00.943] tapdisk_vbd_debug: vhd:/var/run/sr-mount/myZFSraid50/diskVM1.vhd: state: 0x00000000, new: 0x00, pending: 0x06, failed: 0x00, completed: 0x00, last activity: 1758683796.081829, errors: 0x0000, retries: 0x00c2, received: 0x000f9b42, returned: 0x000f9b3c, kicked: 0x0439e111
Sep 24 05:27:00: [05:27:00.943] vhd_debug: /var/run/sr-mount/myZFSraid50/diskVM1.vhd: QUEUED: 0x000871bb, COMPLETED: 0x000871b5, RETURNED: 0x000871b4
Sep 24 05:27:00: [05:27:00.943] vhd_debug: WRITES: 0x000437db, AVG_WRITE_SIZE: 28.221727
Sep 24 05:27:00: [05:27:00.943] vhd_debug: READS: 0x000439e0, AVG_READ_SIZE: 23.678925
Sep 24 05:27:00: [05:27:00.943] vhd_debug: ALLOCATED REQUESTS: (352 total)
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 92: vreq: xenvbd-14-768.67.0, err: 0, op: 2, lsec: 0x02d74e6f, flags: 0, this: 0x1b02ed0, next: (nil), tx: (nil)
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 150: vreq: xenvbd-14-768.67.0, err: 0, op: 2, lsec: 0x01b7f720, flags: 0, this: 0x1b07580, next: (nil), tx: (nil)
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 190: vreq: xenvbd-14-768.67.0, err: 0, op: 2, lsec: 0x01879d78, flags: 0, this: 0x1b0a640, next: (nil), tx: (nil)
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 220: vreq: xenvbd-14-768.67.0, err: 0, op: 2, lsec: 0x02d74ec7, flags: 0, this: 0x1b0cad0, next: (nil), tx: (nil)
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 235: vreq: xenvbd-14-768.67.0, err: 0, op: 2, lsec: 0x02d74ee7, flags: 0, this: 0x1b0dd18, next: (nil), tx: (nil)
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 259: vreq: xenvbd-14-768.67.0, err: 0, op: 2, lsec: 0x18f628e2, flags: 10, this: 0x1b0fa58, next: (nil), tx: 0x1af8a68
Sep 24 05:27:00: [05:27:00.943] vhd_debug: BITMAP CACHE:
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 0: blk: 0x2d74, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1afb228, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 1: blk: 0x34f2b, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1afb080, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 2: blk: 0x1bdb, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1afaed8, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 3: blk: 0xd237, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1afad30, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 4: blk: 0x072f, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1afab88, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 5: blk: 0x0740, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1afa9e0, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 6: blk: 0x8a08, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1afa838, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 7: blk: 0x0773, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1afa690, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 8: blk: 0x3ee8, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1afa4e8, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 9: blk: 0x1b7f, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1afa340, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 10: blk: 0x070e, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1afa198, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 11: blk: 0x11512, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af9ff0, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 12: blk: 0x26ca, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af9e48, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 13: blk: 0xe6f3, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af9ca0, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 14: blk: 0x0fc2, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af9af8, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 15: blk: 0x1c900, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af9950, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 16: blk: 0x96ac, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af97a8, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 17: blk: 0x0790, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af9600, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 18: blk: 0x0734, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af9458, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 19: blk: 0x0752, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af92b0, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 20: blk: 0x0732, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af9108, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 21: blk: 0x1c1bd, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af8f60, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 22: blk: 0xfc4c, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af8db8, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 23: blk: 0x1d0d, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af8c10, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 24: blk: 0x18f62, status: 0x0000000a, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 8, in use: 1, tx: 0x1af8a68, tx_error: 0, started: 1, finished: 1, status: 1, reqs: 0x1b0fa58, nreqs: 1
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 25: blk: 0x1960b, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af88c0, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 26: blk: 0x134b, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af8718, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 27: blk: 0x1bbb, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af8570, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 28: blk: 0x1879, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af83c8, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 29: blk: 0x076a, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af8220, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 30: blk: 0xb73b, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af8078, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: 31: blk: 0x0731, status: 0x00000000, q: (nil), qnum: 0, w: (nil), wnum: 0, locked: 0, in use: 0, tx: 0x1af7ed0, tx_error: 0, started: 0, finished: 0, status: 0, reqs: (nil), nreqs: 0
Sep 24 05:27:00: [05:27:00.943] vhd_debug: BAT: status: 0x00000000, pbw_blk: 0x0000, pbw_off: 0x00000000, tx: (nil)
Sep 24 05:27:00: [05:27:00.943] vhd_debug: /run/sr-mount/myZFSraid50/disk_an_other_VM.vhd: QUEUED: 0x00099bd9, COMPLETED: 0x00099bd9, RETURNED: 0x00099bd9
Sep 24 05:27:00: [05:27:00.943] vhd_debug: WRITES: 0x00000000, AVG_WRITE_SIZE: 0.000000
Sep 24 05:27:00: [05:27:00.943] vhd_debug: READS: 0x00099bd9, AVG_READ_SIZE: 23.188253
Sep 24 05:27:00: [05:27:00.943] vhd_debug: ALLOCATED REQUESTS: (352 total)
Sep 24 05:27:00: [05:27:00.943] vhd_debug: BITMAP CACHE:
Sep 24 05:27:00: [05:27:00.943] vhd_debug: BAT: status: 0x00000000, pbw_blk: 0x0000, pbw_off: 0x00000000, tx: (nil)
Sep 24 05:27:01: [05:27:01.910] watchdog_cleared: vhd:/var/run/sr-mount/myZFSraid50/diskVM1.vhd: watchdog timeout: requests were blocked
Sep 24 06:03:03: [06:03:03.461] tapdisk_vbd_check_progress: vhd:/var/run/sr-mount/myZFSraid50/diskVM1.vhd: watchdog timeout: pending requests idle for 11 seconds
Sep 24 06:03:04: [06:03:04.400] watchdog_cleared: vhd:/var/run/sr-mount/myZFSraid50/diskVM1.vhd: watchdog timeout: requests were blocked
Sep 24 06:10:47: [06:10:47.027] tapdisk_vbd_check_progress: vhd:/var/run/sr-mount/myZFSraid50/diskVM1.vhd: watchdog timeout: pending requests idle for 11 seconds
Sep 24 06:10:48: [06:10:48.416] watchdog_cleared: vhd:/var/run/sr-mount/myZFSraid50/diskVM1.vhd: watchdog timeout: requests were blocked
Sep 24 06:27:16: [06:27:16.740] tapdisk_vbd_check_progress: vhd:/var/run/sr-mount/myZFSraid50/diskVM1.vhd: watchdog timeout: pending requests idle for 100 seconds
Sep 24 06:27:18: [06:27:18.465] watchdog_cleared: vhd:/var/run/sr-mount/myZFSraid50/diskVM1.vhd: watchdog timeout: requests were blocked

                    2025-09-24 08_45_26-Backup — Mozilla Firefox.png

                    Finally, when I say massive backups, my hypervisor is at 50% of its occupancy rate.

                    And this weekend, I'm thinking of skipping all the scheduled backup tasks to be on the safe side.

                    I don't know what to think anymore

                    1 Reply Last reply Reply Quote 0
                    • olivierlambertO Offline
                      olivierlambert Vates 🪐 Co-Founder CEO
                      last edited by

                      Probably worth trying regular local ext SRs (in mdadm if needed) and see how it behave. KISS principle is very strong in general.

                      Y 1 Reply Last reply Reply Quote 0
                      • Y Offline
                        yann_m @olivierlambert
                        last edited by

                        Unfortunatly this is also my conclusion.

                        This is what I used to do with smaller servers.
                        This is the first time with such a "huge server" + hardware RAID + ZFS with dedup.

                        Thanks
                        Have a nice day

                        1 Reply Last reply Reply Quote 0
                        • olivierlambertO Offline
                          olivierlambert Vates 🪐 Co-Founder CEO
                          last edited by

                          It will be a lot better with future SMAPIv3 drivers (it might include ZFS if we want to officially support it)

                          1 Reply Last reply Reply Quote 0
                          • Y Offline
                            yann_m
                            last edited by

                            Do you think the problem could be due to ZFS deduplication ?

                            I just saw that it's possible to disable deduplication "without restarting" .
                            I think my problem is due to excessive disk writes (adding then deleting a snapshot, then coalescing?)
                            Without deduplication, less disk usage ?

                            Or am I wrong ?

                            1 Reply Last reply Reply Quote 0
                            • olivierlambertO Offline
                              olivierlambert Vates 🪐 Co-Founder CEO
                              last edited by olivierlambert

                              In general, at least the consensus, is to NOT use deduplication with ZFS. It's probably one cause of the issue, yes. It's not worth the RAM needed.

                              1 Reply Last reply Reply Quote 0
                              • Y Offline
                                yann_m
                                last edited by

                                Yes, I should probably create a new post ?
                                To discuss and "to survive" : how to make my current server usable again

                                • starting by testing with deduplication turned off.
                                1 Reply Last reply Reply Quote 0
                                • olivierlambertO Offline
                                  olivierlambert Vates 🪐 Co-Founder CEO
                                  last edited by

                                  Let's continue there, you can always rename your thread 🙂

                                  1 Reply Last reply Reply Quote 0
                                  • Y Offline
                                    yann_m
                                    last edited by

                                    Okay, but I can't edit the subject.

                                    1 Reply Last reply Reply Quote 0
                                    • olivierlambertO Offline
                                      olivierlambert Vates 🪐 Co-Founder CEO
                                      last edited by

                                      You can edit your first post, which contains the title

                                      Y 1 Reply Last reply Reply Quote 0
                                      • Y Offline
                                        yann_m @olivierlambert
                                        last edited by

                                        @olivierlambert said in HCP-NG 8.3 freezes and / or reboot on HGR-SAP3 by Ovh:

                                        You can edit your first post, which contains the title

                                        thanks a lot

                                        I'll wait until next week to disable deduplication (never on Friday 😉 )

                                        1 Reply Last reply Reply Quote 1

                                        Hello! It looks like you're interested in this conversation, but you don't have an account yet.

                                        Getting fed up of having to scroll through the same posts each visit? When you register for an account, you'll always come back to exactly where you were before, and choose to be notified of new replies (either via email, or push notification). You'll also be able to save bookmarks and upvote posts to show your appreciation to other community members.

                                        With your input, this post could be even better 💗

                                        Register Login
                                        • First post
                                          Last post