eMMC failed

[Der Geist der Maschine]

My  mmc heavily degenerates as well:

My emmc is from Nov 2019. The dump2fs Life Time writes start Dec 2019, e.g. don't log the usage of the first month. That is negligible.

Code:

2025/02/08: Life Time writes  913GB(*), Life Time Estimate 0x01 (00% - 10%)
2025/04/28: Life Time writes  998GB(*), Life Time Estimate 0x03 (20% - 30%)
2025/05/04: Life Time writes 1000GB(*), Life Time Estimate 0x04 (30% - 40%)

(*) dumpe2fs /dev/dm-0  2> /dev/null | egrep "Lifetime"

Over writing 85GB (50% of these 85GB are for updating Debian 12 to Debian 13), the life time degenerated by 20%. Over writing further 2 GB, the life time degenerated further 10%. This degeneration is not proportional and I expect this emmc to die in the next few days or weeks.

Writing 1000GB is like writing the whole 128GB emmc less than 8 times. That is a very low number. The data-sheet https://files.pine64.org/doc/datasheet/p...0v1.13.pdf suggests that the expected life time is writing 88,0000 GB for my 128GB emmc module (given a certain workload mix).

A questions to others. What the health of your emmcs: Run this command:

Code:

mmc extcsd read /dev/mmcblk2 | grep Life

I shut down my Pinebook Pro. Logout;

[Dendrocalamus64]

My PBP is still on a brand new Hardkernel 256GB emmc so it's 0x01.

My Rockpro64 has a 128GB emmc from Pine64. The filesystem on that dates from August 2021, lifetime writes 1409 GB.
eMMC Life Time Estimation A [EXT_CSD_DEVICE_LIFE_TIME_EST_TYP_A]: 0x02
eMMC Life Time Estimation B [EXT_CSD_DEVICE_LIFE_TIME_EST_TYP_B]: 0x01
eMMC Pre EOL information [EXT_CSD_PRE_EOL_INFO]: 0x01

https://www.cnx-software.com/2019/08/16/...sh-memory/

The Extended Device Specific Data, formerly named Extended Card Specific Data, is where health reports are made available. It contains:

a vendor proprietary health report, 32 bytes long
device lifetime estimation type A, providing health status in increments of 10%
This refers to SLC blocks in our eMMC.
device life time estimation type B, providing health status in increments of 10%
This refers to MLC blocks in our eMMC.
pre-EOL info, reflecting device lifetime by average reserved blocks
returns values normal, warning (80% of reserved blocks consumed) and urgent (90% of reserved blocks consumed)

I don't think the one I had fail was from writing too much, I think it was a bug in the on-emmc firmware that corrupted the internal data structures leading to sudden failure. There does not have to be any warning, data on emmcs needs to be kept backed up at all times.

Although I hadn't set up zram yet back then and was swapping to the emmc, now that I have, iostat -h shows me that the kB_wrtn/s for zram0 is less than for mmcblk2, 9.9k vs 14.5k. With my usage, swapping to the emmc would less than double the data written.

This is with the following config changes applied:
https://wiki.archlinux.org/title/Zram#Op...ap_on_zram

IE, a higher swappiness (180) than the default (60).

However, vm.watermark_boost_factor = 0 prevents periodic "swap storms".
I don't have any data on how much that increased usage. I guess I could run with the default again to collect it.

The default value of watermark_boost_factor is terrible, it's what causes the PBP to lock up in out-of-memory conditions instead of having the OOM killer actually work.

Fwd: Debian 11: Tuning kernel parameters swappiness and watermark_boost_factor to stop SWAP Storm

[Der Geist der Maschine]

My PBP is still on a brand new Hardkernel 256GB emmc so it's 0x01.

I need a new eMMC an noticed the Pine Store is not shipping to my country in these days.

My understanding is that eMMC sockets are not standarized. I noticed Hardkernel offers multiple eMMCs. How do I know which one fit into the Pinebook Pro? Are there also other manufacturers?

My Rockpro64 has a 128GB emmc from Pine64. The filesystem on that dates from August 2021, lifetime writes 1409 GB.

eMMC Life Time Estimation A [EXT_CSD_DEVICE_LIFE_TIME_EST_TYP_A]: 0x02
eMMC Life Time Estimation B [EXT_CSD_DEVICE_LIFE_TIME_EST_TYP_B]: 0x01
eMMC Pre EOL information [EXT_CSD_PRE_EOL_INFO]: 0x01

That Pine-branded eMMC is expected to support writing 88TB. Maybe it's all more complicated, but that tells me that it should degenerate by 10% every 8.8TB writes. Your eMMC is there already after 1.4TB. Good luck.

[Dendrocalamus64]

My understanding is that eMMC sockets are not standarized. I noticed Hardkernel offers multiple eMMCs. How do I know which one fit into the Pinebook Pro? Are there also other manufacturers?

I didn't find any other integrators. Kingston makes the actual emmc and Hardkernel puts it on a board.

https://www.hardkernel.com/?s=256gb&post...ct&lang=en

I think these are all the same socket, which is an unoffical standard, just pre-flashed with different OSes for different devices.

You need USB3.0 eMMC Module Writer 2 : Works with ODROID Orange, Red and Blue eMMC modules

(It is not compatible with the Black eMMC Modules)

I got the orange dot version, with no OS preinstalled, and it fit.

Their wiki article reveals that "orange, red, blue, black" refers to the PCB color, not the color of the marker on the chip.

https://wiki.odroid.com/accessory/emmc/reference_chart

So all the ones currently listed in the shop are the same PCB version with the same connector. Specifically I got the orange PCB (v0.5) with the orange dot (no OS). There is no orange dot on the actual chip.

eMMC connector specification

The connector is made by LS-Mtron Korea.
On the eMMC module, we've used the GB042-34S-H10 (Socket-34pin).
On the host board, we've used the GB042-34P-H10 (Plug-34pin).

The connector specification is here

eMMC connector reliability

The manufacturer says their guaranteed durability of the connector is 30 Cycle of inserting and separating actions.

We need to be careful we don't hit that, it's lower than expected.

That Pine-branded eMMC is expected to support writing 88TB. Maybe it's all more complicated, but that tells me that it should degenerate by 10% every 8.8TB writes. Your eMMC is there already after 1.4TB. Good luck.

After four years of usage. I'll replace it due to obsolescence, i.e. storage requirements increasing and larger modules becoming available, long before it fails (in another 36 years?)

I've seen people on other forums running at 0x0b (110% used up) with no actual effects. I don't think the life time estimation works very well.

and [edit] again:
Most of the storage is configured as MLC - multilevel cell eraseblocks. So the type B life estimation is the one we should be looking at. 0x01 indicates 0 - 10%. So after 1.4 TB, it is still somewhere within 0 - 10% device life time used, which is even within spec.