timeout on meteonews causes ALL after to die **solved**

[lrosenman]

on both 4.9za and 4.9zb, I was seeing a timeout on the meteonet process reporting to meteonews.

When I run it with -d, I shows the timeout, and then "Alarm Clock" and then process exit.

Can this be changed to log the error, and continue on?

[admin]

"-d" is a one time invocation. When running in daemon mode it should not stop.
Do you have evidence that is stops operation in daemon mode? You can check with "rcmeteonet status"

[lrosenman]

Yes. I got whines from CWOP, WeatherBug, et al that they were NOT receiving data.

When I ran it without -d, I got a "Alarm Clock" not, and process exit.

I ran it with -d to figure out what was happening.

Please look at your timeout handling for meteonews.

[lrosenman]

and:
root@raspberrypi:~# rcmeteonet status
-bash: rcmeteonet: command not found
root@raspberrypi:~# ls /etc/init.d/
README kbd nfs-common single
bacula-fd keymap.sh ntp skeleton
bootlogs killprocs procps ssh
bootmisc.sh kmod rc thttpd
checkfs.sh loggerd rc.local udev
checkroot-bootclean.sh meteonetd rcS udev-mtab
checkroot.sh motd reboot umountfs
cron mountall-bootclean.sh rmnologin umountnfs.sh
dbus mountall.sh rpcbind umountroot
ddclient mountdevsubfs.sh rsync urandom
exim4 mountkernfs.sh rsyncd watchdog
gpm mountnfs-bootclean.sh rsyslog wd_keepalive
halt mountnfs.sh samba wear
hostname.sh mtab.sh screen-cleanup x11-common
hwclock.sh networking sendsigs
root@raspberrypi:~#

so, where is this rcmeteonet command.

BTW, this error was on the 5 minute CRON invocation of meteonet.

root@raspberrypi:~# crontab -l | grep meteonet
*/5 * * * * sleep 120; /home/meteohub/meteonet
root@raspberrypi:~#

[lrosenman]

the rapid-fire daemon WAS sending data.

[admin]

Yes. I got whines from CWOP, WeatherBug, et al that they were NOT receiving data.

When I ran it without -d, I got a "Alarm Clock" not, and process exit.

I ran it with -d to figure out what was happening.

Please look at your timeout handling for meteonews.

That is neither an evidence nor does a timeout there aborts anything from a code perspective.
Please check with "rcmeteonetd status" if process is still there or check "ps ax" output
for a process called "meteonet". If the process really drops, then please give me access to your
rig on port 22 so I can check what goes on.

[lrosenman]

As I said, the -r daemon is fine.

the cron one was dieing on the meteonews timeout.

I disabled meteonews, and cwop, et al started getting data again.

How can I get you ip/creds/etc to look at my box?

[lrosenman]

root@raspberrypi:~# ps ax|grep meteonet
2488 ? S 24:07 /home/meteohub/meteonet -r
12300 ? Ss 0:00 /bin/sh -c sleep 120; /home/meteohub/meteonet
12559 pts/0 S+ 0:00 grep meteonet
root@raspberrypi:~#

root@raspberrypi:~# rcmeteonetd status
Usage: /etc/init.d/meteonetd {start|stop|reload|restart|force-reload}
root@raspberrypi:~#

I *AM* an experienced Linux/Unix/etc sysadmin.

[lrosenman]

and as I said, the rapidfire *WAS* working.

[lrosenman]

look at the findu.com history for EW3654 (my station) from Friday to Sunday. You will see **NO** data until I removed Meteonews.

[admin]

I *AM* an experienced Linux/Unix/etc sysadmin.

Will you grant me access or do you think you can fix it yourself?

[lrosenman]

I will grant you access, but want to pass credentials, and ip/port privately.

[lrosenman]

PM with creds sent.

[admin]

I installed an updated version on your rig that better handles the overall timeout.
Will also be part of the next general update.

[lrosenman]

Thank You.