Bug #15126
closedSG-1100 pfSense+ recovery results in non aligned disk slices
100%
Description
Currently preparing for an upgrade of SG-1100 remote worker fleet.
However after installing the latest SG-1100 recovery image (pfSense-plus-compat-recovery-23.09.1-RELEASE-aarch64.img.gz) it appears that the resulting image restore to SG-1100 eMMC is not aligned:
(reference https://docs.netgate.com/pfsense/en/latest/solutions/sg-1100/reinstall-pfsense.html)
gpart show mmcsd0 => 1 15273599 mmcsd0 MBR (7.3G) 1 409600 1 efi (200M) 409601 131072 2 fat32 (64M) 540673 14732927 3 freebsd [active] (7.0G)
This is a UFS build. Clearly the FreeBSD slice (starting sector 540673) is not aligned with 4k / 32k / 1M boundary. Non aligned writes may have an impact on eMMC life (depends on write workload of course).
Within the slice the actual UFS partition is at least 8k aligned (although suboptimal given that the UFS2 default block size is 32kB):
gpart show mmcsd0s3 => 0 14732927 mmcsd0s3 BSD (7.0G) 0 16 - free - (8.0K) 16 14732911 1 freebsd-ufs (7.0G)
Compare this to a Netgate 7100 (with ZFS):
gpart show mmcsd0 40 61071280 mmcsd0 GPT (29G) 40 1024 1 freebsd-boot (512K) 1064 984 - free - (492K) 2048 4194304 2 freebsd-swap (2.0G) 4196352 56872960 3 freebsd-zfs (27G) 61069312 2008 - free - (1.0M)
Hopefully image build can be corrected using appropriate gpart add -t freebsd -a 1M ... /dev/mmcsd0
argument parameters.
Lastly is the SG-1100 (aarch64) recovery image also used for SG-2100? If so this issue may also impact SG-2100.