Recent content by sigo

Do you plan to update 5.11 kernel in pve 6.xx version to latest ZFS (2.0.7)? User-space tools was updated to 2.0.7 recently.

For my PM983 SSD NVMe (zfs mirror) I have approx. 15000 fync/sec.

I will try to test pvetest version with fio and report here.

ZFS 0.8.4 now in the pve-no-subscription repo. Does it contain same patches as in pvetest version? Specially - patches from arenas to speedup zvol.

yes, this solves the bug. Thanks!

I will check this fix later and report here.

Latest update to PVE broke my containers. I had installed a simplified Debian containers from LinuxContainer.org (for example: https://us.images.linuxcontainers.org/images/debian/buster/amd64/default/20191119_05:24/rootfs.tar.xz) and have a "bind mount" to ZFS directory in lxc configuration...

what error you get when innodb_doublewrite set to off? and show "zfs get all rpool/....." for you mysql filesystem

I had the same problem. This happened in the EX- series, but when I switched to the PX -series, the error went away.

Why aren't you using MSSQL for Linux? ( https://docs.microsoft.com/en-us/sql/linux/quickstart-install-connect-ubuntu?view=sql-server-2017) You can install mssql4linux in LXC and have direct access to ZFS pools (via bind mount). You can achieve a high performance by this way.

I finally had the opportunity to check the questions - problem disappear after removing 99-default.link and reappear again if this file (99-default.link) stored in /etc/systemd/network/ Checked on latest Proxmox 6.0 with 5.0.18 kernel.

Which distribution used inside LXC?

This is production server, I will can try to check your suggestions later.

root@pixie:~# cat /etc/udev/udev.conf # see udev.conf(5) for details # # udevd is started in the initramfs, so when this file is modified the # initramfs should be rebuilt. #udev_log="info" root@pixie:~# ls -alFRh /etc/udev/ /etc/udev/: total 15K drwxr-xr-x 4 root root 5 Apr 9 20:41 ./...

I checked the old logs (kern.log and syslog) with previous bootings (correct and failed, these files are attached), and it seems that after updating the packages the kernel(systemd) stopped renaming eth0 (default detected device name) into a consistent name.