meteohub.de

Hi Boris,

I'm getting an no response sometimes from the \"Inspect Data\" webpage on my hub.

Here are the errors. Any ideas?

Mar 14 23:16:24 (none) daemon.info thttpd[1755]: spawned CGI process 4933 for file 'cgi-bin/meteohub.cgi'
Mar 14 23:16:54 (none) daemon.err thttpd[1755]: killed CGI process 4933
Mar 14 23:17:00 (none) cron.info /USR/SBIN/CRON[4978]: (root) CMD (sleep 10; /srv/www/meteograph.cgi graph Today-CT2 >/data/graphs/Today-CT2.png; /srv/www/cgi-bin/stamp.cgi mv /data/graphs/Today-CT2.png /data/uploads/Today-CT2.png)
Mar 14 23:17:01 (none) cron.info /USR/SBIN/CRON[4982]: (root) CMD (/bin/cat /proc/uptime >/home/meteohub/alive)
Mar 14 23:17:01 (none) cron.info /USR/SBIN/CRON[4980]: (root) CMD (/home/meteohub/wswincut 2>>/data/log/messages)
Mar 14 23:17:18 (none) syslog.info -- MARK --
Mar 14 23:17:24 (none) daemon.info thttpd[1755]: spawned CGI process 5009 for file 'cgi-bin/meteohub.cgi'
Mar 14 23:17:54 (none) daemon.err thttpd[1755]: killed CGI process 5009

time outs do happen from time to time when the system is extremely busy. If it happens again please switch to the system information pages and have a look at the first system load value. If it is larger than 5 or so, the system really has a severe load. I have seen loads above 10 and then the system gets very unresponsive to the web interface... but still performs fine in data logging and graph generation.

It happens almost ever time I try to go into that menu.

My load just before I clicked \"Inspect Data\" was: 2.86, 2.42, 2.59

After it failed my load was: 2.89, 2.52, 2.60

It happens 9 out of 10 tries.

This screen is what Meteohub was doing when it failed. Not sure if it will help.


Mem: 28460K used, 1916K free, 0K shrd, 224K buff, 17880K cached
Load average: 2.47 2.47 2.55 (Status: S=sleeping R=running, W=waiting)
PID USER STATUS RSS PPID %CPU %MEM COMMAND
18556 root R N 564 18215 93.6 1.8 wmr928eval
1042 root S 716 1 2.8 2.3 meteonet
18551 root R 872 18537 1.4 2.8 top
1060 root S 2844 1059 1.1 9.3 wmr928d
2 root SWN 0 1 0.4 0.0 ksoftirqd/0
18524 root S 2032 1007 0.2 6.6 sshd
172 root SW 0.0 1.7 udevd
1097 root S 516 1 0.0 1.6 getty
18193 root S 468 1070 0.0 1.5 cron
18582 root S 440 1070 0.0 1.4 cron
956 root S 420 1 0.0 1.3 watchdog
1062 root S 368 1 0.0 1.2 rsync
912 daemon S 356 1 0.0 1.1 portmap
1059 root S 296 1 0.0 0.9 wmr928d


Even when the load is only \"1.90, 1.78, 1.81\" I can't get to this screen.

Hi,
just checked my systems. not problem there. Anybody else having symptoms like this?

Your system load looks ok. Is it just the \"inspect data\" page or are other pages also impacted?

So far only the 'Inspect Data' page times out when asked to load.

On average I notice that most menu pages take 3-6 seconds to load into my browser. Is this normal?

The 'System Info' page takes about 7-8 seconds to load.

How do these times compare to others?

Reason was an extreme amount of data at the end of the month caused by the vantage logging all sensors every couple of seconds. raw data has become >70 MB in March.

Solution introduced with 2.2d strips redundant data down and so reduces total amount of raw data to go though in order to find the actual position in the raw data file.

Increased priority for web server processes introduced with 2.2e should make Meteohub's web interaction even more snappy.