It was just last week that Linux optimizations were leading to possible 6M IOPS per core and then at the start of this week new patches pushed Linux past 7M IOPS per-core with an ideal hardware configuration as well. In ending out the week, 8M IOPS has been reached!

Linux block subsystem maintainer and lead IO_uring developer Jens Axboe had a goal of hitting 7M IOPS per-core performance this week. On Monday he managed to already hit 7.2M IOPS and today hit 7.4M IOPS with his latest work-in-progress kernel patches.

Linux block subsystem maintainer and IO_uring lead developer Jens Axboe continues making staggering optimizations to the kernel code to squeeze out the maximum performance potential out of his shiny new system.

The EROFS read-only file-system is adding LZMA compression support using a new header/container format called MicroLZMA.

In April 2020 there was the somewhat surprising announcement of Micron announcing their own open-source storage engine designed for SSDs and persistent memory. The Heterogeneous-Memory Storage Engine (HSE) has been redesigned and on Monday debuted in v2.0 form with some fundamental changes.

Linux kernel developers have been working tirelessly to squeeze more performance out of IO_uring and the block / I/O code in general. IO_uring lead developer Jens Axboe who also serves as the Linux block subsystem's maintainer (among other roles and major contributions over the years) has used his system as a baseline for evaluating such kernel improvements. He's now moved to using AMD Zen 3 while sticking to Intel Optane storage and is seeing a mighty speed boost out of AMD's latest processors.

It's been a while since having any news to report on Bcachefs as the promising open-source file-system born out of the Linux kernel's block cache code. However, Kent Overstreet continues working tirelessly on it and has now merged Bcachefs' snapshot support.

It was just last month when ~3.5M IOPS per-core was impressive with the code for Linux 5.15 to further push Linux's I/O limits. Now for code likely to be included in Linux 5.16 it's currently at 3.8M IOPS with a single tread.

While the OpenZFS 2.1 feature release has been available since July, for those still using the OpenZFS 2.0.x series and not yet prepared to make the jump to that big new release with dRAID and other changes, OpenZFS 2.0.6 was released this week.

While there is now KSMBD with Linux 5.15 for offering an in-kernel SMB file server, its scope is much more limited than that of the Samba project in user-space. With that said, Samba 4.15 is out now with its latest batch of features and improvements for open-source SMB/CIFS support on Linux and other platforms.

Following the big OpenZFS 2.1 release from July that brought Distributed SPARE RAID, a compatibility property for pools, and other new features, OpenZFS 2.1.1 is available today as a follow-up release for this open-source ZFS file-system implementation for Linux and FreeBSD systems.

OverlayFS continues to be used by Linux IoT/embedded devices and other use-cases as a union mount file-system. With Linux 5.15 the OverlayFS file-system continues to improve.

With Linux 5.15 there are optimizations for EXT4, big improvements for XFS, and significant work on Btrfs too. Rounding out the notable file-system work on Linux 5.15, the F2FS updates were submitted and subsequently landed for this next kernel version.

While Samsung has explicitly stated before that queued TRIM works for Samsung 860 SSDs on Linux and thus leading to only older Samsung 840/850 drives being blocked from queued TRIM usage, that turns out to be inaccurate and now more quirks are added for the Samsung 860 and 870 series SSDs on Linux.

It looks like Paragon Software's NTFS3 kernel driver providing much better Linux support for the Microsoft NTFS file-system will land for the 5.15 kernel!

While Linux 5.15 has added KSMBD as an SMB3 in-kernel file server, the NFS code within the Linux kernel continues advancing as well for network file sharing needs. With NFS in Linux 5.15 are a few notable improvements.

It's busy on the Linux file-system front for the 5.15 cycle with Btrfs adding a degenerate RAID option along with performance improvements to big improvements for XFS and now comes the EXT4 updates.

In addition to Btrfs sporting new features with Linux 5.15, the XFS and EROFS file-systems also have some shiny new features and improvements for this next version of the Linux kernel.

The Btrfs file-system updates have landed now in Linux 5.15 mainline with some exciting new features and improvements.

One of the earliest pull requests sent in for the now-open Linux 5.15 cycle was proposing KSMBD land as the in-kernel SMB3 file server as an alternative on Linux systems to running Samba in user-space. At the time it wasn't clear if Linus Torvalds would pull in this file server code to the Linux kernel but now he has indeed landed it.

One of the very first pull requests for Linux 5.15 now that its merge window is open following the Linux 5.14 release is to merge KSMBD, the in-kernel SMB3 protocol file server.

A number of Phoronix readers have been asking about some fresh file-system comparisons on recent kernels. With not having the time to conduct the usual kernel version vs. file-system comparison, here are some fresh benchmarks looking at the Btrfs, EXT4, F2FS, and XFS file-system benchmarks on a speedy WD_BLACK SN850 NVMe solid-state drive.

Announced back in April of 2020 was Micron's HSE as an open-source storage engine optimized for solid-state drives and persistent memory. After quickly crossing the v1.0 milestone and working its way up to HSE 1.9 at the end of last year, there hasn't been any updates since. Fortunately, the project is still alive and HSE 2.0 is on the way.

While Linux 5.14 stable is expected next weekend, the Reiser4 (and experimental Reiser5) file-system driver code has been finally updated for compatibility with the Linux 5.13 kernel series.

Merged to the mainline Linux kernel six years ago was the LightNVM subsystem as part of Linux 4.4 LTS around "Open-Channel SSDs". That LightNVM code is now slated for removal with the upcoming Linux 5.15 cycle.

Introduced to the Linux kernel earlier this year with Linux 5.12 was IDMAPPED mounts that allow for different mounts to expose the same file or directory with different ownership. IDMAPPED mounts was designed with use-cases ranging from containers to systemd-homed and more as outlined in the earlier article. Btrfs is now ready to begin supporting IDMAPPED mounts.

Those having to deal with data stored on NTFS partitions from Linux have been eagerly awaiting the "NTFS3" kernel driver that Paragon Software has been working now for a year to upstream into the Linux kernel. No pull request has been sent in yet but the twenty-seventh spin of this driver was published today.

Last month I wrote about a possible global counter for block/disk changes on Linux being discussed by Microsoft and systemd developers to better track changes via a system-wide monotonically increasing number as an alternative to the existing per-disk tracking. That functionality is now queued up as part of the block subsystem changes ahead of the Linux 5.15 merge window in a few weeks.

A lot of new code landed yesterday in the Btrfs file-system's "for-next" Git branch ahead of the Linux 5.15 merge window opening up in about one month's time.

The NVMe specification provides for an abrupt shutdown mode over the normal/safe shutdown command if needing to quickly get the NVMe solid-state storage ready for powering off as quickly as possible. Currently the Linux kernel isn't making use of the NVMe abrupt shutdown command but a proposal by Micron is looking to begin its usage.