On Sun, 06 Jan 2019 13:43:26 +0000 Martin Gregorie wrote:

> On Sun, 06 Jan 2019 10:23:19 +0000, Markus Robert Kessler wrote:
>
>> Therefore I have 3 user processes running:
>>
>> - one process under root, writing to /dev/watchdog every 5 seconds -
>> one ping to Fritzbox every 1 minute and writing to >> /tmp/fritzlog -
>> one ping to net printer, 1/min. and writing to >> /tmp/printerlogfile
>>
> Did you write these processes yourself or are they from a package? The
> reason for asking is that, if they are your code, consider reversing the
> logging logic and only reporting problems to the log, rather than what
> sound like endless 'I'm OK' messages. Apart from reducing wear on the SD
> card, it will make the logs a lot easier to read.


The code is selfmade.

And, yes, also successful pings are reported. So, there are 3 things that
can happen: online, offline, neither / error.

As far as I've learned about SSHDs, SSD cards, USB memory sticks is, that
the built-in memory controller takes care of never writing the same
memory cells over and over -- and risk early defects.

Instead the formerly used cells are marked as erased / to be re-used, and
the new, modified content of the file will be written to a different
location. So, all cells are stressed equally.

This seems not to be the case here.


>> I am / was using KINGSTON 8 GB cards, installed with Raspbian stretch
>> out of the box, installed straight forward with boot and root
>> partition.
>> Root fs partition had > 2 GB available after install.
>>
> I used to use Kingston years ago, stopped when I found out that they are
> only a rebrander, i.e. that their SD cards no long come from a single,
> known foundry.


OK, that makes things clearer...


> I now use Sandisk, because they make their own product, for everything:
> cameras, RPi as well as the flight logger, navigation PNA and FLARM
> system in my glider. The latter two are safety-critical systems.
>
> I'm not quoting usage figures for my RPi because its run time is just an
> hour or two per week.
>
> The only problem I have with it, an early 512MB Model B, is that if I
> leave it idle for an hour or two after booting it appears to be dead
> when I try to log in to it over SSH. However, if I log in and out as
> soon as it has come up after booting then its still responsive a couple
> of hours later.


See second branch


>> The system was running for about 2 week and I had access to all the
>> data on both partitions, but I needed to do a reboot, and I now had to
>> find out, that the Root fs partition (/dev/sdc2 when trying to mount on
>> a normal PC) is completely unaccessible. Plugging in shows up both
>> partitions, but every attempt to mount will fail.
>>
> Obvious question: is the PC running Linux? If not, then it will only be
> able to read the boot partition.


Yes, it's Linux, and after installing there are 2 partitions offered for
mounting.

But after the crash / failure, when I clicked on the "root fs" partition,
there only popped up an error alert saying that this partition cannot be
mounted. Same symptoms seen when the first Kingston crashed.

And, after running a fsck / e2fsck now, there is no such partition any
more. I can access the whole partition like 'cat /dev/sdc2 | xxd | less',
but I only see the first 64 MB. The remaining space seems unavailable.
Aaaargh...

Best regards,

Markus


--
Please reply to group only.
For private email please use http://www.dipl-ing-kessler.de/email.htm

--- SoupGate-Win32 v1.05