The fact that this does not work on some systems, means what you said: systemd seems to be confused somehow and does something different than we want. Not sure if we can control that
No worries at all, Dan! I realize and appreciate all your efforts. When pi-hole is down you realize how nasty the unfiltered web can be!
I'm wondering if you have a base system you use - eg which release of the pi image you develop with (eg raspian jessie? Which release?). I'm willing to start totally from a new install if it means we can make this work.
I found it necessary to do an uninstall and reinstall moving to using FTL.
Just to add to that: I suspect in the end part of the problem was a browser caching issue. Even after uninstalling and reinstalling, the dashboard screen was not rendering properly. After I blew my cache in Chrome away everything rendered properly.
I generated another debug log - even though my FTL seems to be running since last night with no new changes. Funny the debug log generation was stuck on parsing the lighttpd/error.log for nearly 30 minutes (on a Pi3). My Pi is based on a Raspian Jessie full installation, if I recall.
Token: wtzkw0itvm
Although the FTL was working at the time, I also tried:
pi@raspberrypi:~ $ sudo netstat -tulpn | grep FTL
tcp 0 0 127.0.0.1:4711 0.0.0.0:* LISTEN 721/pihole-FTL
pi@raspberrypi:~ $ pihole-FTL running
FATAL: Opening of FTL log (/var/log/pihole-FTL.log) failed!
Make sure it exists and is writeable by user pi
pi@raspberrypi:~ $ sudo chown pi:pi /var/log/pihole-FTL.log
pi@raspberrypi:~ $ pihole-FTL running
Yes: Found a running FTL process
Thanks for the note on the debugger. We've had another one or two reports of lockups. That code hasn't really changed, so will have to tear it apart to see what's going on.
FTL is going down for me each day ... seems to be tied to the nightly log purge? I run the purge at 23:58 (instead of midnight because I have chronometer e-mail me some stats first). Notice there's nothing in the pihole-FTL.log after 23:58 until I restarted it at 8:44am this morning:
[2017-05-03 23:58:10.581] NOTICE: pihole.log has been flushed
[2017-05-03 23:58:10.581] Resetting internal data structure
[2017-05-03 23:58:10.582] Queries in memory before flushing: 24144
[2017-05-04 08:44:42.538] ########## FTL started! ##########
[2017-05-04 08:44:42.538] FTL branch: (no branch)
[2017-05-04 08:44:42.538] FTL hash: v2.6
[2017-05-04 08:44:42.539] FTL date: 2017-04-22 16:15:35 +0200
[2017-05-04 08:44:42.539] FTL user: pi
[2017-05-04 08:44:42.540] Warning: Starting pihole-FTL directly is not recommended.
[2017-05-04 08:44:42.540] Instead, use system commands for starting pihole-FTL as service (systemctl / service).
[2017-05-04 08:44:42.540] Notice: Found no readable FTL config file
[2017-05-04 08:44:42.541] Using default settings
Not sure how that can happen. FTL is constructed such that it should catch all errors and write them into the logfile before quitting. The only exception to this is if it gets killed by SIGKILL (because I cannot catch that by construction).
I suggest to use this instructions to couple the debugger to FTL:
It should catch even SIGKILL and even if not but the pihole-FTL crashes, we are in the best position to do debugging.
Not sure if it helps, but my Primary PiHole box is on Dedicated HW, that one fails. I have a Backup running on a Virtual Machine and it has been stable. Both systems were rock solid prior to the latest updates
i didnt know if i should post a new topic or just add on to this one. FTL keeps crashing for me as well. i am running in in a virtual machine. OS is ubuntu 16.04. i updated 2 days ago to 3.0, and thats when it started. then i updated to 3.0.2, and it continued. i figured, ok, it was a decent update, i'll completely reinstall and start fresh. didn't help, it is still crashing. also, when the ftl is running for those few minuts after a restart, the 'query types over time' and
'Forward Destinations over Time' do not work. they start working once ftl crashing. . the debug token generated: jlno4oda3v