webserver becomes non-responsive over time.

[fireheadman]

What's Wrong

Have noticed after running garage_pi for several weeks it will eventually stop loading the webpage.

Environment Info

RPi Model: rpi2

OS Release: jessie

Service Status:

Attach Logs

[fireheadman]

I have attempted the following:

• checked for full filesystems, everything was within tolerance (nothing full or close to full)
• remove all logs in the garage_pi/instance directory, then reboot.... still no webserver working.
• removed the history.db, still not webserver working

The only thing that worked was to restore a previous backup that is known to be good.... then reboot and my webserver works.

So there must be something else, perhaps a caching tolerance that reaches a limit? Not sure where to check this.

[SWiT]

You mentioned purging the logs. Was the sd card full when you did? I didn't have the magnetic reed switch wired up yet. So the input was "dangling" randomly switching the status between open and closed. This was generating log files very quickly. I could see that filling up the storage after a few days or a week or two depending on size.

[LachyTech]

I have seen similar behaviour as well, but I find that I can't ssh into the device either when it is happening so I need to do a hard reset. When I log back in, I notice that the history has continued to be populated with open/close events even when it seemed unresponsive. I am thinking it could also be network/wifi related as it doesn't appear in my router's DHCP client list. Are you having any network issues @fireheadman?

The SD card is nowhere near full in my case.

pi@garagepi:~ $ df -h
Filesystem      Size  Used Avail Use% Mounted on
/dev/root       7.0G  2.3G  4.4G  35% /
devtmpfs         87M     0   87M   0% /dev
tmpfs           215M     0  215M   0% /dev/shm
tmpfs            86M  696K   86M   1% /run
tmpfs           5.0M  4.0K  5.0M   1% /run/lock
/dev/mmcblk0p1  253M   49M  204M  20% /boot
tmpfs            43M     0   43M   0% /run/user/1000

[LachyTech]

I've had this running successfully for over two weeks now, the cause for my issue was that wifi was dropping and not reconnecting automatically. I applied the following fix to /etc/wpa_supplicant/wpa_supplicant-wlan0.conf from this article: https://unix.stackexchange.com/questions/566992/embedded-device-wpa-supplicant-reconnect

# Enable WiFi scanning on network disconnect.
ap_scan=1
# Make wpa_supplicant do the scan instead of relying on the driver.
disable_scan_offload=1