Firecracker

Infrastructure & DevOpsApache-2.0

Lightweight microVMs for serverless and container workloads.

Latest v1.18.0 · by Amazon Web ServicesWritten in RustWebsitefirecracker-microvm/firecrackerRSS

Release activity

Release activity — 11 releases across 9 days since Jan 20, 2026. Each cell is one day; darker means more releases that day. Nothing is recorded before Jan 20, 2026. Older weeks are hidden at this screen width.
AugSepOctNov
SundayNo releases on Aug 2, 2026No releases on Aug 9, 2026No releases on Aug 16, 2026No releases on Aug 23, 2026No releases on Aug 30, 2026No releases on Sep 6, 2026No releases on Sep 13, 2026No releases on Sep 20, 2026No releases on Sep 27, 2026No releases on Oct 4, 2026No releases on Oct 11, 2026No releases on Oct 18, 2026No releases on Oct 25, 2026No releases on Nov 1, 2026No releases on Nov 8, 2026No releases on Nov 15, 2026
MondayNo releases on Aug 3, 2026No releases on Aug 10, 2026No releases on Aug 17, 2026No releases on Aug 24, 2026No releases on Aug 31, 2026No releases on Sep 7, 2026No releases on Sep 14, 2026No releases on Sep 21, 2026No releases on Sep 28, 2026No releases on Oct 5, 2026No releases on Oct 12, 2026No releases on Oct 19, 2026No releases on Oct 26, 2026No releases on Nov 2, 20261 release on Nov 9, 2026
TuesdayNo releases on Aug 4, 2026No releases on Aug 11, 2026No releases on Aug 18, 2026No releases on Aug 25, 2026No releases on Sep 1, 2026No releases on Sep 8, 2026No releases on Sep 15, 2026No releases on Sep 22, 2026No releases on Sep 29, 2026No releases on Oct 6, 2026No releases on Oct 13, 2026No releases on Oct 20, 2026No releases on Oct 27, 2026No releases on Nov 3, 2026No releases on Nov 10, 2026
WednesdayNo releases on Aug 5, 2026No releases on Aug 12, 2026No releases on Aug 19, 2026No releases on Aug 26, 2026No releases on Sep 2, 2026No releases on Sep 9, 2026No releases on Sep 16, 2026No releases on Sep 23, 2026No releases on Sep 30, 2026No releases on Oct 7, 2026No releases on Oct 14, 2026No releases on Oct 21, 2026No releases on Oct 28, 2026No releases on Nov 4, 2026No releases on Nov 11, 2026
ThursdayNo releases on Aug 6, 2026No releases on Aug 13, 2026No releases on Aug 20, 2026No releases on Aug 27, 2026No releases on Sep 3, 20262 releases on Sep 10, 2026No releases on Sep 17, 2026No releases on Sep 24, 2026No releases on Oct 1, 2026No releases on Oct 8, 2026No releases on Oct 15, 2026No releases on Oct 22, 2026No releases on Oct 29, 2026No releases on Nov 5, 2026No releases on Nov 12, 2026
FridayNo releases on Aug 7, 2026No releases on Aug 14, 2026No releases on Aug 21, 2026No releases on Aug 28, 2026No releases on Sep 4, 2026No releases on Sep 11, 2026No releases on Sep 18, 2026No releases on Sep 25, 2026No releases on Oct 2, 2026No releases on Oct 9, 2026No releases on Oct 16, 2026No releases on Oct 23, 2026No releases on Oct 30, 2026No releases on Nov 6, 2026No releases on Nov 13, 2026
SaturdayNo releases on Aug 8, 2026No releases on Aug 15, 2026No releases on Aug 22, 2026No releases on Aug 29, 2026No releases on Sep 5, 2026No releases on Sep 12, 2026No releases on Sep 19, 2026No releases on Sep 26, 2026No releases on Oct 3, 2026No releases on Oct 10, 2026No releases on Oct 17, 2026No releases on Oct 24, 2026No releases on Oct 31, 2026No releases on Nov 7, 2026No releases on Nov 14, 2026

11 releases since Jan 20, 2026, busiest day 2

Branches

1.18
v1.18.0
1.17
v1.17.0
1.16
v1.16.2
1.15
v1.15.1
1.14
v1.14.4

Changelog

Filter releases by branch
11 of 11 releases
v1.18.0Latest

Firecracker v1.18.0

Added 3
  • Added support for Nitro Sandboxes, allowing Firecracker to launch and manage isolated VMs carved out of the parent EC2 instance by the Nitro Hypervisor as an alternative to KVM
  • Firecracker drives the Nitro Enclaves driver at /dev/nitro_enclaves instead of /dev/kvm for Nitro Sandboxes while keeping its existing API, jailer and virtio device model
  • Added the PUT /enclave API to configure a Nitro Sandbox in confidential or direct mode

From Firecracker

Added
  • #6393: Added support for Nitro Sandboxes. Firecracker can now launch and manage isolated VMs that the Nitro Hypervisor carves out of the parent EC2 instance, as an alternative to KVM. Firecracker drives the Nitro Enclaves driver (/dev/nitro_enclaves) instead of /dev/kvm, and keeps its existing API, jailer and virtio device model. A sandbox is configured with the new PUT /enclave API, in confidential or direct mode. See the Nitro Sandboxes documentation.
View originalPermalink
How v1.18.0 went
v1.17.0

Firecracker v1.17.0

Added 11
  • Added support for virtio device reset.
  • Added opt-in metrics fields set via PUT /metrics or a config file: emit_id emits the microVM instance id under a top-level id field, and properties emits operator-defined key-value pairs under a top-level properties field, each independent.
  • Added a Transparent option for the huge_pages setting that uses transparent huge pages for guest memory via madvise(MADV_HUGEPAGE); guest memory must be a multiple of 2MB when using this option.
  • Added official support for 6.18 microVM guest kernels.
  • Added support for booting bzImage guest kernels on x86_64 in addition to uncompressed ELF (vmlinux) images, with the kernel image format detected automatically.
  • Added the huge_pages field to PUT /snapshot/load, allowing the restored microVM to reuse the snapshot's host page configuration or select None, Transparent, or 2M.
Changed 2
  • Changed the T2S CPU template to set the FB_CLEAR bit of MSR_IA32_ARCH_CAPABILITIES to 1, letting guest kernels recognize that VERW clears fill buffers, including on host kernels before v6.4.
  • Gated the CLIRD_EL1 override so it only applies on host kernels 6.10 or newer, alleviating a guest performance regression seen on host kernels in the range 6.3 to 6.10.
Fixed 5
  • Fixed the vsock device permanently suppressing host-to-guest RX delivery after a bare pause/resume cycle without a snapshot, by persisting the TRANSPORT_RESET RX gate in device state and having the resume kick respect it.
  • Fixed virtio-mem leaving unplugged memory writable by the VMM, and potentially unmapped, on microVMs restored from a snapshot memory file; each slot is now mapped with the correct protection using MAP_NORESERVE and the process aborts if re-mapping fails.
  • Fixed virtio-mem discarding the whole hotpluggable region on every UNPLUG_ALL request, now skipping the discard when the range has no plugged blocks.
  • Fixed virtio-mem block accounting becoming inconsistent with KVM memory slots when a plug or unplug request failed partway through, by committing each slot's state only after its KVM update succeeds.
  • Fixed a TOCTOU race in the aarch64 jailer when setting ownership of the CPU cache and MIDR_EL1 information files copied into the chroot.
View originalRead the notes
How v1.17.0 went
v1.16.2

Firecracker v1.16.2

Changed 3
  • Gate CLIRD_EL1 override to only happen on host kernels equal or newer than 6.10 release, alleviating the guest performance regression seen on host kernels in range [6.3..6.10) correlated to incorrect cpu cache topology presented to the guest
  • Bumped the snapshot format version so snapshots created by this version cannot be loaded by 1.16.0 or 1.16.1, and snapshots created by those versions cannot be loaded by this one
  • On single-vcpu microVMs on the newest Intel hosts (m7i, m8i), host-to-guest vsock throughput may decrease by ~15% (median) because the single guest vCPU rather than the host becomes the bottleneck
Fixed 10
  • Fixed the vsock device permanently suppressing RX (host-to-guest) delivery after a bare pause/resume cycle, where the resume kick armed the TRANSPORT_RESET RX gate without a reset event having been sent; the gate is now part of the persisted device state and the resume kick respects it
  • Fixed virtio-mem leaving unplugged memory writable by the VMM, and potentially unmapped, on microVMs restored from a snapshot memory file by mapping each slot with the correct protection and the MAP_NORESERVE flag, aborting if the re-map fails
  • Fixed virtio-mem discarding the whole hotpluggable region on every UNPLUG_ALL request, even when nothing was plugged; the discard is now skipped when the range has no plugged blocks
  • Fixed virtio-mem leaving its block accounting inconsistent with the KVM memory slots when a plug or unplug request failed part way through, by committing each slot's state only after its KVM update succeeds
  • Fixed a TOCTOU race in the aarch64 jailer when setting ownership of the CPU cache and MIDR_EL1 information files copied into the chroot
  • Fixed memory hotplug sizes silently wrapping when converted from MiB to bytes by bounding requested_size_mib on PATCH /hotplug/memory and total_size_mib, block_size_mib and slot_size_mib on PUT /hotplug/memory to 32 bits, so oversized values are rejected instead of becoming a 0-byte region
View originalRead the notes
How v1.16.2 went
v1.16.1

Firecracker v1.16.1

Fixed 2
  • Reverted the use of O_NOFOLLOW for the jailer's cgroup and network namespace file operations, so symlinks are again allowed in these paths.
  • Fixed a bug that caused vsock guest-to-host connections to time out after snapshot restore when a snapshot was taken with a TX descriptor in-flight, by replaying the TX queue notification on restore so in-flight TX descriptors are re-processed and notification suppression is re-armed.
View originalRead the notes
How v1.16.1 went
v1.16.0

Firecracker v1.16.0

Added 9
  • Developer preview support for hotplugging and hot-unplugging PCI virtio devices (block, pmem, net) on a running microVM
  • Support for Vsock Unix domain socket path overriding on snapshot restore
  • Optional rate limiting for serial console output, configurable via the rate_limiter field on PUT /serial, with a new uart rate_limiter_dropped_bytes metric
  • Per-callsite rate limiting for error, warn, and info log messages, with a warn-level summary of suppressed messages and a new rate_limited_log_count metric
  • Rate-limiter support for the virtio-pmem device to control I/O bandwidth generated by FLUSH requests from the guest
  • Notification suppression support in the virtio-vsock device via the EVENT_IDX virtio feature
Fixed 11
  • Capped virtio-rng per-request entropy to 64 KiB to prevent excessive host memory allocation from a guest-constructed descriptor chain
  • Corrected the HID of the VMGenID device to align with the upstream Linux kernel so the driver binds correctly before Linux 6.10
  • Fixed guest UART driver getting stuck and stopping transmission after snapshot restore when a snapshot was taken during serial transmission
  • Fixed missing /sys/devices/system/cpu/cpu*/cache/* in aarch64 guests on host kernels >= 6.3 with guest kernels >= 6.1.156
  • Fixed virtio-mem plug/unplug skipping KVM slot updates for memory blocks not aligned to a slot boundary
  • Bounded balloon statistics descriptor length to prevent a guest-controlled oversized descriptor from stalling the VMM event loop
View originalRead the notes
How v1.16.0 went
v1.15.1

Firecracker v1.15.1

Added 1
  • Added clock_realtime flag to the LoadSnapshot API to let users using kvm-clock explicitly opt in to advancing the clock with KVM_CLOCK_REALTIME
Fixed 5
  • Cap virtio-rng per-request entropy to 64 KiB so a guest cannot make Firecracker allocate more host memory than the guest provided
  • Fixed missing /sys/devices/system/cpu/cpu*/cache/* in aarch64 guests when running on host kernels >= 6.3 with guest kernels >= 6.1.156
  • Fixed virtio-mem plug/unplug skipping KVM slot updates for memory blocks not aligned to a slot boundary, which could leave hotplugged memory inaccessible to the guest or leave freed memory accessible to it
  • Bound balloon statistics descriptor length so a guest-controlled oversized descriptor cannot stall the VMM event loop, affecting microVMs with stats_polling_interval_s > 0
  • Fixed monotonic clock jumping on restore by the wall-clock time elapsed since the snapshot for x86_64 guests using the kvm-clock clock source on host Linux >= 5.16
Security 2
  • Enforce the virtio device initialization sequence in the PCI transport, matching MMIO transport behavior, including status transition validation, queue configuration and feature negotiation restrictions, blocking re-initialization after a failed reset, and setting DEVICE_NEEDS_RESET when device activation fails (fixes CVE-2026-5747)
  • Reject device status writes that clear previously set bits in the MMIO transport, except for reset
View originalRead the notes
How v1.15.1 went
v1.14.4

Firecracker v1.14.4

Added 1
  • Added clock_realtime flag to the LoadSnapshot API so users of kvm-clock can explicitly opt in to advancing the clock with KVM_CLOCK_REALTIME
Fixed 4
  • Cap virtio-rng per-request entropy to 64 KiB so a guest cannot make Firecracker allocate more host memory than the guest provided
  • Fixed missing /sys/devices/system/cpu/cpu*/cache/* in aarch64 guests when running on host kernels >= 6.3 with guest kernels >= 6.1.156
  • Fixed virtio-mem plug/unplug skipping KVM slot updates for memory blocks not aligned to a slot boundary, which could leave hotplugged memory inaccessible to the guest or leave freed memory accessible to the guest
  • Fixed monotonic clock jumping on snapshot restore by the wall-clock elapsed time for x86_64 guests using the kvm-clock clock source on host Linux >= 5.16
Security 3
  • Enforce the virtio device initialization sequence in the PCI transport, matching the MMIO transport, including status transition validation, queue configuration window checks, feature negotiation state checks, blocking re-initialization after a failed reset, and setting DEVICE_NEEDS_RESET when device activation fails (CVE-2026-5747)
  • Reject device status writes that clear previously set bits in the MMIO transport, except for reset
  • Bound balloon statistics descriptor length so a guest-controlled oversized descriptor cannot stall the VMM event loop, affecting only microVMs with stats_polling_interval_s > 0
View originalRead the notes
How v1.14.4 went
v1.14.3

Firecracker v1.14.3

Fixed 1
  • Fixed validation of TCP SYN options length when MMDS is enabled
View originalRead the notes
How v1.14.3 went
v1.15.0

Firecracker v1.15.0

Added 2
  • Add support for the VMClock device, implementing the snapshot safety features but providing no clock-specific information yet for guest clock synchronization
  • Added Intel Granite Rapids as a supported and tested platform for Firecracker on 6.1 host kernel versions
Changed 2
  • VMClock support uses one extra GSI for the VMClock device, reducing available GSIs for VirtIO devices; new maximum is 92 devices on Aarch64 and 17 devices on x86
  • Jailer binary copy process disallows symlinks and hardlinks at the destination path and changes ownership of the copied binary to the specified uid/gid
Fixed 4
  • Fixed possible ENXIO error during file open operation when the underlying file is a FIFO without active readers already attached
  • Fixed vsock local port reuse across snapshot restore by saving the last used local port into the snapshot, requiring users to regenerate snapshots
  • Fixed a bug that corrupted memory files of differential snapshots for VMs with multiple memory slots, affecting VMs using memory hot-plugging or x86 VMs larger than 3GiB
  • Fixed validation of TCP SYN options length when MMDS is enabled
View originalRead the notes
How v1.15.0 went
v1.14.2

Firecracker v1.14.2

Fixed 2
  • Fixed a possible ENXIO error that could occur during file open operation if the underlying file is a FIFO without active readers already attached.
  • Fixed a bug that caused Firecracker to corrupt the memory files of differential snapshots for VMs with multiple memory slots, affecting VMs using memory hot-plugging or any x86 VMs with a memory size larger than 3GiB.
View originalRead the notes
How v1.14.2 went
v1.14.1

Firecracker v1.14.1

Changed 1
  • Update binary copy process inside Jailer to disallow symlinks and hardlinks at the destination path and change ownership of the copied binary to the specified uid/gid
View originalRead the notes
How v1.14.1 went
View all

Discussion

If you publish Firecracker, you can claim this product by proving you administer its repository.