meteohub.de

Hi skyewright
Thanks for Information.... by the way: can I get back to 4.2e like upgrading to a higher version?

Greetings from sunny Germany!

The wrong checksums are repeating exactly every 30 minutes...

Rikken wrote:The wrong checksums are repeating exactly every 30 minutes...

Possibly a conflict with some other radio device and the two things happen to try to tramsmit at the same moment every 30 minutes (but the transmissions between those occasions are fine)?

As some of you reporting problems with recomputation my thought is that recomputation gets aborted by something and therefore data is not build up completely. As my rig does not show this effect it would be great help, if one of you who can reproduce the effect on his system would give me access to allow me some debugging.

@Rikken: Checksum error during logging data from wmrs-200 is not related to these troubles. I would consider that as usual behavior. It looks like one of the bytes for checksum has not been transferred for any reason. Dropping a packet once in half an hour would not worry me. If you are at NSLU2 and many other tasks are running in parallel (recomputation for example) this might happen.

Yes, transmissions between are ok. But I think that the problem is not another radio device, cause never had wrong checksums with earlier versions. But who knows?
By the way: The station seems to work fine even with the wrong checksums, all data are displayed ok.

Hi Boris

you might be right, cause my recomputation is still running. So i will wait, till its finished.
Thanks for information!

regards

It will probably help Boris if the community can answer the following question:

Is this related to a specifc sensor type?
Or
Is this related to a specific weather station model.

Just list the weather station and the sensor model that shows the error. Maybe we are able to narrow the context of this issue in this way.

Hi,

In my case the problem does not seem to be related to any specific sensor. "hist-xxx" files under the weather directory are incomplete and finish in the middle of different sensor records, depending on the particular file.
I am using OS WMR-200

Boris,

Can you please confirm weather it is safe to downgrad from version 4.2f to version 4.2e while the problem is solved?

Thanks

admin wrote:if one of you who can reproduce the effect on his system would give me access to allow me some debugging.

You have my details. You are welcome to visit. If you take a look at hist-day1 you'll find that the file is around 8KB and stops at:

day1_th5_dewmin_c 6

(well it did when I just looked, I think that when I looked earlier it had stopped at a different point?).

That's on meteohub-c running 4.2f.

meteohub-b running 4.2e and recording less sensors has a hist-day1 that is ~13KB and finishes at:

day1_thb9_sealevelmax_inhg 30.14

Downgrade to 4.2e is considered save. Just install 4.2e update.

skyewright wrote:You have my details. You are welcome to visit. If you take a look at hist-day1 you'll find that the file is around 8KB and stops at:

day1_th5_dewmin_c 6

Quick update on this:
All the hist files that refresh each 5 minutes seem to be affected; all finishing 'short' but at various places.
At each refresh the hist-day1 file usually stops at around the point quoted above, but not always exactly there.

I found a bug that leads to segfaults (you can see these in "dmesg" logfile related to wmr928eval.

Brand new update 4.2g should fix this. You can update the leazy way with "Check Update" feature on "Maintenance" page.

admin wrote:I found a bug that leads to segfaults (you can see these in "dmesg" logfile related to wmr928eval.

Brand new update 4.2g should fix this. You can update the leazy way with "Check Update" feature on "Maintenance" page.

Check Update worked great, but...

4.2g installed. Post-update recomputation completed. The hist files are all exactly as before I installed 4.2g - i.e. they have not been recalculated either during the recomputation nor since (and its been more than 5 minutes since completion).

dmesg reports segfaults...

Not to complicate things, but I'm running 4.2g and with my computation completed I'm NOT seeing problems anymore since the upgrade. All of my history appears to be displaying properly now.