[
admin wrote:hat is the experience of our "early x86 adopters"?
As mentioned in other threads I'm now running an ALIX 1D.
I'll not repeat what I've mentioned elsewhere, but in essence the install went smoothly.
At present the ALIX is in Demo mode, and the NSLU2 is still running as before. After I've had the ALIX running for a few days I hope to move the licence over.
The ALIX is being fed exactly the same data as the NSLU2 by having it connected to NSLU2 using the TCP/IP meteohub server option.
Earlier today I did a Meteohub Application Data backup on the NSLU2 and restored that on the ALIX.
The backup/restore went smoothly (restore was
much faster than backup) but inevitably it left the ALIX with a short gap in the data between the end of the backup and the time of restore. I plugged that gap by getting both SAMBA shares visible on another computer, then I stopped logging on the ALIX, quickly copied the raw file for this month from NSLU2 to ALIX, then restarted logging and initiated a Recompute. Now the two Meteohubs have exactly the same data except for a couple of anemometer readings (and it was flat calm at that time anyway). :cheer:
An alternative would have been to wait January then copy or patch the 200812 raw without stopping logging?
Recompute on the ALIX is
much faster! :cheer: A recompute on the NSLU2 yesterday took 6 hours or so. I haven't timed a recompute on the ALIX (it might be nice if start and end of recompute appeared in the meteohub log, as happens with backups?) but it's a lot less - I did a couple of full recomputes this afternoon in less time than one would take on the NSLU2!
Looking at html files generated from the same template on both Meteohubs, I seem to see the Month and Year values updating a little sooner on the ALIX, but maybe that's just a coincidence?
One strange thing, and maybe Boris can comment on this is that the size of
exactly the same data on the ALIX is substantially less!
NSLU2
ALIX 1D
Maybe some file system corruption or 'dead' files on the NSLU2? Or some thing else?