Partition wiped after power loss on WP8548

sometimes power loss occur, and sometimes my WP8548 starting with wiped partition with all user data loss.

Here attached a chunk of core log which I suppose somehow related to this problem:

mount: mounting /dev/ubi2_0 on /mnt/flash failed: Invalid argument
S02mount_early: Unable to mount /dev/ubiblock2_0 onto /mnt/flash.
[    9.822066] block ubiblock2_0: released
[    9.824965] ubi2: detaching mtd4
[    9.863207] ubi2: mtd4 is detached
[    9.968411] yaffs: dev is 32505860 name is "mtdblock4" rw
[    9.973233] yaffs: passed flags "tags-ecc-off"
S02mount_early: YAFFS2 file system on user1 partition is empty, forcing it to be UBIFS.
Erasing 128 Kibyte @ 77a0000 -- 100 % complete 
ubiformat: mtd4 (nand), size 125566976 bytes (119.8 MiB), 958 eraseblocks of 131072 bytes (128.0 KiB), min. I/O size 2048 bytes
libscan: scanning eraseblock 957 -- 100 % complete  
ubiformat: 958 eraseblocks are supposedly empty
ubiformat: formatting eraseblock 930 -- 97 % comp[   23.245872] ubi2: attaching mtd4
ubiformat: formatting eraseblock 957 -- 100 % complete  

Is there any workaround to this? I don’t want my data to be wiped.

I run my custom legato image based on Legato linux source.

Any help will be appreciated.

+1, it happened to me too.

Was that with Release 13.1? Or some other version?

I am using custom-built image based on latest Legato Linux Source Distribution Version 16.10.3

I am using the same version.

Any updates on this topic?

Looks like this ends up corrupting the ubifs partition and it ends up wiping it (as per design) to go back to a functional system.

Was the target doing any operations when this occurs or is it just idling or maybe even sleeping?
I guess you could force a sync to limit the occurrence in the meantime.

Yes, target is active and performing some I/O operations on the partition during runtime, I assume it fails when data is writing to the partition and sudden power loss occur, I can accept this particular data to be lost, but not entire partition.