SD card health monitoring

emonSD
1

For over a year now we’ve been including sandisk industrial grade SD cards in all emonPi, emonBase and emonSD units. We’ve not had a single reported issue.

As mentioned on the bottom of this thread, thanks to @wjcroft it’s possible to monitor the health of these cards:

This post is to document how to do this:

Connect via SSH then download sdmon:

curl -L https://github.com/Ognian/sdmon/releases/download/latest/sdmon-armv7.tar.gz | tar zxf -

then run sdmon:

sudo ./sdmon /dev/mmcblk0

this should give you an output like this

"version": "v0.4.1-3 (63a79fd) armv7",
"date": "2022-08-11T02:05:07.000Z",
"device":"/dev/mmcblk0",
"addTime": "false",
"signature":"0x44 0x53",
"SanDisk":"true",
"manufactureYYMMDD": "210712",
"healthStatusPercentUsed": 1,
"featureRevision": "0x1f",
"generationIdentifier": 5,
"productString": "SanDisk                         ",
"success":true
}

This particular card which has been in heavy use for over a year is at 99% health.

In future releases of emonSD we could look at making this metric display on the Emoncms admin page.

sdmon can only read health data from certain industrial SD cards:

Here’s the sdmon repo:

2

Thanks @glyn.hudson

Out of interest, this is a raspberry Pi that’s been running for over 5 years continuously (not my emonPi unfortunately). Certainly looking like I need to prepare a replacement:

pi@raspberrypi:~ $ sudo ./sdmon /dev/mmcblk0
{
"version": "v0.4.1-3 (63a79fd) armv7",
"date": "2022-08-14T21:35:03.000Z",
"device":"/dev/mmcblk0",
"addTime": "false",
"read_via_cmd56_arg_1":"read successful but signature 0xff 0xff",
"idata.response[]":"0x900 0x00 0x00 0x00",
"flashId": ["0x45","0x3a","0x98","0xb3","0x76","0x72","0x0d","0x0e","0x00"],
"icVersion": ["0x1f","0xc3"],
"fwVersion": [38,131],
"ceNumber": "0x01",
"spareBlockCount": 41,
"initialBadBlockCount": 9,
"goodBlockRatePercent": 99.40,
"totalEraseCount": 425285,
"enduranceRemainLifePercent": 37.20,
"avgEraseCount": 628,
"minEraseCount": 563,
"maxEraseCount": 684,
"powerUpCount": 127,
"abnormalPowerOffCount": 47,
"totalRefreshCount": 0,
"productMarker": ["0x00","0x00","0x00","0x00","0x00","0x00","0x00","0x00"],
"laterBadBlockCount": 0,
"success":true
}
3

This really needs a wrapper script to post the output to MQTT and an systemd timer file to trigger it once in x (or CRON but I’m adverse to CRON). Pull in the MQTT server details from an ini rather than edit the script.

@glyn.hudson would be a great addition to the emonscripts install (optional).

I don’t have a card this will read unfortunately.

I also think it would be worth adding to the Upgrade & Update part of the guide a recommended list of SD Cards.

4

Interesting, thanks for sharing. Do you know what SD card this is from? What version of emonSD is it running? The output looks different from our current Sandisk SDSQQNR-032G-GN6IA

5

I don’t actually know what card is in that Pi - I’ll check next time I get to it.

Like I say, it’s not from my emonPi, it’s an ADSB receiver in my loft - been running for 5+ years. A few power outages here and there, and pretty nasty temps in summer. I just found it interesting for something that’s been running for such a long time to see the real-world stats on it

6

Here’s the results from a brand new Kingston 32GB “High Endurance” card:

pi@emonpi:~ $ sudo ./sdmon /dev/mmcblk0
{
"version": "v0.4.1-3 (63a79fd) armv7",
"date": "2022-09-03T16:01:27.000Z",
"device":"/dev/mmcblk0",
"addTime": "false",
"read_via_cmd56_arg_1":"read successful but signature 0xff 0xff",
"idata.response[]":"0x900 0x00 0x00 0x00",
"flashId": ["0x98","0x3a","0x98","0xb3","0x76","0x72","0x08","0x0e","0x00"],
"icVersion": ["0x20","0x25"],
"fwVersion": [29,63],
"ceNumber": "0x02",
"spareBlockCount": 9,
"initialBadBlockCount": 10,
"goodBlockRatePercent": 99.67,
"totalEraseCount": 694,
"enduranceRemainLifePercent": 99.97,
"avgEraseCount": 1,
"minEraseCount": 1,
"maxEraseCount": 1,
"powerUpCount": 11,
"abnormalPowerOffCount": 0,
"totalRefreshCount": 0,
"productMarker": ["0x00","0x00","0x00","0x00","0x00","0x00","0x00","0x00"],
"laterBadBlockCount": 0,
"success":true
}

I’ve swapped to a new card following all of the Dire Warnings about card life. To be fair, the old one was that well known brand “LD”.