Cannot access Meteohub through webpage **solved**

[cbhiii]

Hi,

I think my Meteohub is getting too busy to respond to my requests. I tried many times tonight to access it through the webpage and it would not respond. It did not respond earlier today but I thought it was my connection at work.

My log file shows that the process is being killed after a certain amount of time.

Mar 24 20:46:53 (none) daemon.info thttpd[6291]: spawned CGI process 23805 for file 'cgi-bin/meteohub.cgi'
Mar 24 20:47:00 (none) cron.info /USR/SBIN/CRON[23821]: (root) CMD (/bin/cat /proc/uptime >/home/meteohub/alive)
Mar 24 20:47:00 (none) cron.info /USR/SBIN/CRON[23823]: (root) CMD (/home/meteohub/wswincut 2>>/data/log/messages)
Mar 24 20:47:00 (none) cron.info /USR/SBIN/CRON[23822]: (root) CMD (sleep 10; /srv/www/meteograph.cgi graph Today-CT2 >/data/graphs/Today-CT2.png; /srv/www/cgi-bin/stamp.cgi cp /data/graphs/Today-CT2.png /data/myweb/uploads/Today-CT2.png; /srv/www/cgi-bin/stamp.cgi mv /data/graphs/Today-CT2.png /data/uploads/Today-CT2.png)
Mar 24 20:47:23 (none) daemon.err thttpd[6291]: killed CGI process 23805

[cbhiii]

A reboot fixed this problem. Is it possible that over time the unit becomes less responsive due to a bug and a reboot will fix that?

[admin]

when the unit is doing a lot of things and system load goes beyond 6 or so then it can be that a http-request is killed after 30 secs. I can set the timeout threshold up to 60 secs...

the WD and WSWIN value computation do eat some serious processing time. After a reboot things start uncrowded, but you also purged some jobs that have to started later anyhow.

I have not noticed that performance is degrading over time. there are just some peek times in system load (jobs scheduled by cron), where the web interface receives time outs here and there...

hopefully,
you are not creating 30 graphs each second ;-)

[cbhiii]

Can I kill off the WD and WSWIN processes? Are they just for external programs that I do not run? I don't use these programs.

[admin]

Hi,
as I just found out, your performance problems are caused by your NSLU2 just running on 133 MHz instead of 266 MHz as the recent units. You can check your self at the \"BogoMIPS\" info on system information page.

I don't have a 133 MHz unit here, but I guess this baby is too weak in performance to do all the magic we need. As 133 MHz units are no longer sold for years, you really got a very very old one.

Good news is that the NSLU2 can be modded quite easily to \"full speed\". Details here.. and as always... you do it at your own risk!

http://www.nslu2-linux.org/wiki/HowTo/OverClockTheSlug