On Sunday, December 25, 2016 at 8:20:16 PM UTC-5, Marc Aubry wrote:
>
> hi
> i though i got it last night after having removed the ip v6 part of the 
> rpi and hardcoding the ip address of the mysql server, the weewx did not 
> crash for 12 hours. but this morning at 07:12 the weewx had crashed again 
> the last line of the syslog regarding weather station is :
>
> *Dec 26 07:12:32 weatherstation weewx[602]: wh23xx: read_record: cmdbuf: 
> 02 02 04 04 (len=4)*
>
> would it suggest a wh23xx driver going in loop ?
>

was weewxd still running after 07:12:32?  could you see it when you do 'ps 
aux'?

the 'cmdbuf' log message is normal - the driver sends a command to the 
console, then the console replies.

if the driver were in an infinite loop, the driver would log it.

if the driver were timing out waiting for a reply via usb, the driver would 
log it.

if the console replies with a bogus or unexpected response, the driver 
would log it.

there is a while loop at line 630 that could theoretically result in an 
infinite loop, but if that were happening, there would be log messages 
leading up to and during it.

does any other process write to the log file after 07:12:32?  if not, then 
you probably have a system-wide problem, not specific to weewx.

you say that weewx stops responding frequently.  is weewxd still running 
each time?  does any other process write to the log after the last weewxd 
log entry?  is the last weewxd log entry always the same?

 

> i have no idea how to make the weewx startup script depend on the network 
> and ntp. please could you help me on this one....
>

if you are using an LSB init.d rc script, then you want something like this:

# Required-Start:    $local_fs $remote_fs $syslog $ntp $network

if you are using a systemd unit file, then you want something like this:

Requires=time-sync.target network.target
After=time-sync.target network.target

but that might be redundant - maybe some systemd guru could chime in...

remember - this is not the root cause of your problem - as shown in the 
log, weewx sleeps for two minutes, and by the time it tries a mysql 
connection again, the network and time are squared away.

 

> The log is not stopped is just that nothing else happen on the machine. 
> this rpi is dedicated only to weewx it's a headless machine connected to 
> the lan via ethernet. the only service loaded in top of weewx is tightvnc 
> server. it's like weewx goes in loop somewhere and stop doing anything. 
> this morning i just restart the weewx service and everything went back on 
> normal.
>

does anything else write to the syslog?

try running weewx with the simulator as the driver instead of wh23xx.  see 
if that results in the same behavior.

 

> FYI my wh2310 weather station is a Misol clone of TP2700WC
>

 just to be clear, fine offset manufactures the stations.  they are 
rebranded (not cloned) by various resellers.  depending on the vendor, the 
WH2300 console appears to be packaged with one of three different types of 
instrument clusters - an all-in-one, a two-part, or a three-part cluster.

http://foshk.com/

http://foshk.com/Weather_Professional/WH2300.html

since the wh23xx driver talks to the console, the instrument cluster does 
not matter as far as the driver is concerned.

m

-- 
You received this message because you are subscribed to the Google Groups 
"weewx-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to