Subject line says it all... I run Pi-hole 5 on a Pogoplug V2 (armel, CPU: Feroceon 88FR131 [56251311] revision 1 (ARMv5TE)).
It would be great to be able to test Beta 6 on the armel platform.
Thanks for considering it.
Dave
I'm afraid this will not happen. The people from pi-hole changed their build infrastructure to alpine and alpine does not support this architecture. I have the exact same setup like you have and I build pihole-FTL myself. This works. This should get you going -> docs/docs/ftldns/compile.md at release/v6.0 · pi-hole/docs · GitHub
Also I've seen other people here that are also running pihole on pogo-plugs, so feel free to search this forum.
Please note ... pihole 6 requires a little more memory so you will need to have swap enabled + you have to change the swappiness because otherwise your pihole will be terrible slow. I will check the swappiness setting that works for me when I get home and let you know.
Good luck ... and thanks for keeping the plugs alive <3
2 Likes
Thanks for the response... and I'll give the build process a go.
I replied in the other thread before seeing this one. I use an SSD and as said, definitely use a swap file. It takes a little over 2 hours to compile each time. I've found that anything other than an SSD takes forever. I used a flash drive and a regular spinning hard drive and those took at least 8 or 10 hours if i remember correctly.
1 Like
Oops 
Luckily we have below for that 
1 Like
Using chroot and qemu-user-static, I compiled the deps and binary on my Intel 64bit workstation (amd64 as Debian calls it) in Bookworm. Whole build takes about 20 min IIRC.
Getting everything to happen consistently and reproducibly took a few tries, but got that bashed into submission.
Upgrading at first felt messy... but I changed my build process to include a version number and build-time-stamp on the binary's filename (pihole-FTL-v5.25.2-2341-g91ea8d49-2024-12-15-163830). So, I keep of copy of the v5.25.2 binary in /usr/bin also, and just use a symlink ln -s target-name name to point to it. That way if I accidently run pihole -up, recovery is easy.
Also made sure to do a pihole checkout for the development branch of core, web and ftl. When ftl development checkout was done, I just went in and created the symlink to point to my newly-compiled armel binary and all was good after a pihole restartdns. (I'm thinking that it also complained about setupVars.conf, which I copied from migration_backup_v6/ ... )
Afterthought: If upgrading to V6, don't forget to disable lighttpd...otherwise pihole might try serving http on :8080 ... I'm thinking that this may have happened to me once, and I had to fiddle with pihole-FTL --config [key] to nudge it back to :80.
Note that this is on a non-OXNAS pogoplug. In any case, this gives me some hope for ARMv5TE/armel boxes for at least through the end of support for Debian 13/Trixie. We may hit EOL at some point after that.
2 Likes
Are you still able to run above one?
Mine doesnt anymore after an upgrade yesterday.
Bash completion (double tab):
$ sudo pihole
allow checkout flush regex uninstall
allow-regex debug help reloaddns updateGravity
allow-wild deny logging reloadlists updatePihole
api disable query status version
arpflush enable reconfigure tail wildcard
And if I try anyway:
$ sudo pihole restartdns
Usage: pihole [options]
Example: 'pihole allow -h'
Add '-h' after specific commands for more information on usage
Domain Options:
allow, allowlist Allow domain(s)
deny, denylist Deny domain(s)
--regex, regex Regex deny domains(s)
--allow-regex Regex allow domains(s)
--wild, wildcard Wildcard deny domain(s)
--allow-wild Wildcard allow domain(s)
Add '-h' for more info on allow/deny usage
Debugging Options:
-d, debug Start a debugging session
Add '-c' or '--check-database' to include a Pi-hole database integrity check
Add '-a' to automatically upload the log to tricorder.pi-hole.net
-f, flush Flush the Pi-hole log
-r, reconfigure Reconfigure or Repair Pi-hole subsystems
-t, tail [arg] View the live output of the Pi-hole log.
Add an optional argument to filter the log
(regular expressions are supported)
api <endpoint> Query the Pi-hole API at <endpoint>
Options:
setpassword [pwd] Set the password for the web interface
Without optional argument, password is read interactively.
When specifying a password directly, enclose it in single quotes.
-g, updateGravity Update the list of ad-serving domains
-h, --help, help Show this help dialog
-l, logging Specify whether the Pi-hole log should be used
Add '-h' for more info on logging usage
-q, query Query the adlists for a specified domain
Add '-h' for more info on query usage
-up, updatePihole Update Pi-hole subsystems
Add '--check-only' to exit script before update is performed.
-v, version Show installed versions of Pi-hole, Web Interface & FTL
uninstall Uninstall Pi-hole from your system
status Display the running status of Pi-hole subsystems
enable Enable Pi-hole subsystems
disable Disable Pi-hole subsystems
Add '-h' for more info on disable usage
reloaddns Update the lists and flush the cache without restarting the DNS server
reloadlists Update the lists WITHOUT flushing the cache or restarting the DNS server
checkout Switch Pi-hole subsystems to a different GitHub branch
Add '-h' for more info on checkout usage
arpflush Flush information stored in Pi-hole's network tables
Below is mine:
$ pihole-FTL --config webserver.port
80,[::]:80,443s,[::]:443s
Your config key for webserver port is correct, as long as lighttpd is not fighting for it also...
If you are using the V6 source for pihole-FTL and it built correctly, and it is running correctly... you shoud be able to browse the admin page at http://ip-of-pihole/admin/login . http://ip-of-pihole/admin will also work (either directly, or after redirecting for login).
The previous url had an admin.php in it… not there anymore - and the new web content is needed from git also.
By the way,
pihole-FTL -vv
Can give details about your binary's build.
I struggled a while when upgrading ... I over-wrote some files accidentally and ended up having to reinstall from scratch a few times. I also messed up permissions a few times... and lost track of setupVar.conf and some other things. Eventually, it did become a smooth upgrade process, but only after few mistakes.
If any of the following aren't right, it won't work...
- lighttpd is still running
- if the new web pages aren't installed correctly
- permissions are borked
- user structure isn't correct
- probably some other stuff I miss saying...
Is that slowness in the web admin interface or in resolving DNS (or both?)
And because I have to "It's Kenny Kenny West Y'all"
You don't need lighttpd in v6. You can uninstall it.
Also, if lighttpd is installed and using port 80, Pi-hole v6 will work, but the installer will change the web server port to 8080. Try http://pihole_IP:8080/admin/.
Well, it won't work as expected... ... one can either disable it w/ systemctl and stop any running instances of it, or remove it.
Yes, I liked the embedded civetweb server ... made me think about civet coffee (...not really my cup of tea, but hey...)
Sorry, not a native English speaker here, might need to reformulate that to: "you might want to change the swappiness"
For me, as you can read here: Pihole v6 on low spec hardware having swap with "default" settings resulted in slow DNS responses (webinterface is also not lightning fast, but works). After some very helpful discussion, I ended up disabling swap and running everything from the limited amount of memory (128M).
This caused the pogoplug to go out of memory every now and then ... so, to mitigate that, I activated swap again and (after some reading) lowered the swappiness. This seems to work for now. Although I am not sure how much longer we will be able to keep these things alive.
If you have anything else running besides below ones (my minbase):
$ pstree -cps
systemd(1)─┬─agetty(372)
├─agetty(567176)
├─cron(321)
├─dbus-daemon(327)
├─pihole-FTL(1880538)─┬─{pihole-FTL}(1880553)
│ ├─{pihole-FTL}(1880554)
│ ├─{pihole-FTL}(1880555)
│ ├─{pihole-FTL}(1880556)
│ ├─{pihole-FTL}(1880557)
│ ├─{pihole-FTL}(1880558)
│ ├─{pihole-FTL}(1880559)
│ ├─{pihole-FTL}(1880650)
│ ├─{pihole-FTL}(1880651)
│ ├─{pihole-FTL}(1904395)
│ ├─{pihole-FTL}(1904396)
│ ├─{pihole-FTL}(1989860)
│ └─{pihole-FTL}(1989861)
├─sshd(385)───sshd(2024355)───sshd(2024361)───bash(2024362)───pstree(2024366)
├─systemd-journal(181)
├─systemd-logind(334)
├─systemd-timesyn(301)───{systemd-timesyn}(319)
└─systemd-udevd(207)
You could try disabling them to free some memory.
I dont need anything else to run Pi-hole.
Also no "Network Mangler" bc that can be done with ifupdown --> /etc/network/interfaces.
$ man debootstrap
[..]
minbase, which only includes required packages and apt
EDIT: x86_64 though but only 123MB used (total minus available):
$ free -h
total used free shared buff/cache available
Mem: 957Mi 123Mi 59Mi 3.3Mi 795Mi 834Mi
Swap: 2.0Gi 1.5Mi 2.0Gi
EDIT2: Oh when gravity runs on Sunday, it'll consume a little more oc
Come to think of it, you could also ditch above systemd-timesyncd.service.
Because:
$ pihole-FTL --config ntp.sync.active
true
EDIT:
$ free
total used free shared buff/cache available
Mem: 980000 130612 58540 3392 812156 849388
Swap: 2097148 1536 2095612
$ sudo systemctl disable --now systemd-timesyncd.service
Removed "/etc/systemd/system/dbus-org.freedesktop.timesync1.service".
Removed "/etc/systemd/system/sysinit.target.wants/systemd-timesyncd.service".
$ free
total used free shared buff/cache available
Mem: 980000 129704 59416 3388 812184 850296
Swap: 2097148 1536 2095612
$ pstree
systemd─┬─2*[agetty]
├─cron
├─dbus-daemon
├─pihole-FTL───13*[{pihole-FTL}]
├─sshd─┬─sshd───sshd───bash
│ └─sshd───sshd───bash───pstree
├─systemd-journal
├─systemd-logind
├─systemd-udevd
└─unbound
Thats almost one MB
This should be a minimal install, but I am also running cloudflared to facilitate dns over https. So this is how my pstree looks like:
root@pogo01:~# pstree
init-+-cloudflared---9*[{cloudflared}]
|-cron
|-dbus-daemon
|-getty
|-haveged
|-klogd
|-pihole-FTL---11*[{pihole-FTL}]
|-smartd
|-sshd---sshd---bash---pstree
|-syslogd
`-systemd-udevd
And this is how free looks like:
root@pogo01:~# free -h
total used free shared buff/cache available
Mem: 112Mi 77Mi 2.7Mi 4.6Mi 42Mi 35Mi
Swap: 1.5Gi 23Mi 1.5Gi
root@pogo01:~#
So its all very tight, but still "workable".
Thanks for the hints.
2 Likes
@kennywest , can I ask you two questions please?
- Are your binaries actually ARM, EABI5 version 1, ie.
root@PiHole1-B388:/# file /bin/echo
/bin/echo: ELF 32-bit LSB pie executable, ARM, EABI5 version 1 (SYSV), dynamically linked, interpreter /lib/ld-linux.so.3, BuildID[sha1]=da886736535b17f3a832ea22669663e9c82adbe8, for GNU/Linux 3.2.0, stripped
I was thinking that the OXNAS Pogo's aren't armel, but maybe it's just the kernel that's different... ?
- Does bohdi's OXNAS rootfs come w/
systemdenabled/turned on by default?
I ask the 2nd question because my failover pihole is also 128MB, like your Pogoplug Pro (but w/ the Kirkwood SoC, not Oxford).
Thanks in advance,
Dave
I did a fresh install of the second pogoplug I have, identical as the one currently running pihole v6. The re-installed one was running pihole v5 (rock solid, never seen it crash).
Here is the output of the commands on the fresh installed one (i.e. with rootfs from bohdi):
root@debian:~# ps 1
PID TTY STAT TIME COMMAND
1 ? Ss 0:02 init [2]
root@debian:~# file /bin/echo
/bin/echo: ELF 32-bit LSB pie executable, ARM, EABI5 version 1 (SYSV), dynamically linked, interpreter /lib/ld-linux.so.3, BuildID[sha1]=2e2eca0f79466cc959c19b3d92ef0a5438cf9182, for GNU/Linux 3.2.0, stripped
root@debian:~# uname -a
Linux debian 5.15.158-oxnas-tld-2 #2 SMP PREEMPT Sun May 12 18:45:05 PDT 2024 armv6l GNU/Linux
root@debian:~#
Output of the first command means it is still booting via sys V init ... I guess 
That’s great - that, together with the OXNAS kernel tag says that its just the kernel that is different- I always thought the V3 and Pro required distinct binaries, but I guess they are compatible in that respect.
Thank you for that.
In one of your previous posts it looked like systemd was indeed running… on iPhone now but will peak at it again when home. …. Ahhh now i see it… I think it was de Hakkelaar that posted pstree -cps which showed systemd running…
Do you need above one?
$ apt-file search bin/haveged
haveged: /usr/sbin/haveged
$ apt-file list haveged
[..]
haveged: /etc/init.d/haveged
haveged: /lib/systemd/system/haveged.service
$ apt show haveged
[..]
Description: Linux entropy source using the HAVEGE algorithm
haveged is a userspace entropy daemon which is not dependent upon the
standard mechanisms for harvesting randomness for the system entropy
pool. This is important in systems with high entropy needs or limited
user interaction (e.g. headless servers).
.
haveged uses HAVEGE (HArdware Volatile Entropy Gathering and Expansion)
to maintain a 1M pool of random bytes used to fill /dev/random
whenever the supply of random bits in dev/random falls below the low
water mark of the device.
.
More information about HAVEGE is available at
http://www.irisa.fr/caps/projects/hipsor/
$ apt rdepends haveged
haveged
Reverse Depends:
Suggests: overlayroot
Suggests: shadowsocks-libev
Depends: octavia-agent
Recommends: freedombox
Suggests: fbx-all
And above one sounds a bit redundant if syslogd is already active?
$ apt-file search bin/klogd
busybox-syslogd: /sbin/klogd
$ apt-file list busybox-syslogd
[..]
busybox-syslogd: /etc/init.d/busybox-klogd
busybox-syslogd: /etc/init.d/busybox-syslogd
$ apt show busybox-syslogd
[..]
Description: Provides syslogd and klogd using busybox
The system log daemon is responsible for providing logging of
messages received from programs and facilities on the local host as
well as from remote hosts.
.
The kernel log daemon listens to kernel message sources and is
responsible for prioritizing and processing operating system
messages.
.
The busybox implementation of the syslogd is particular useful on
embedded, diskless (netboot) or flash disk based systems because it
can use a fixed size ring buffer for logging instead of saving logs
to the disk or sending it to remote logging servers. The ring buffer
can be read using the (also busybox based) command logread.
.
This package provides the glue to the busybox syslogd and klogd to be
used in the system by providing the appropriate symbolic links and
scripts.
Do those Pogo's boot from an SD card?
If so, you cant read out SMART from an SD card so above one is redundant too:
$ apt-file search bin/smartd
smartdns: /usr/sbin/smartdns
smartmontools: /usr/sbin/smartd
$ apt-file list smartmontools
[..]
smartmontools: /etc/init.d/smartmontools
smartmontools: /lib/systemd/system/smartmontools.service
$ apt show smartmontools
[..]
Description: control and monitor storage systems using S.M.A.R.T.
The smartmontools package contains two utility programs (smartctl and smartd)
to control and monitor storage systems using the Self-Monitoring, Analysis and
Reporting Technology System (S.M.A.R.T.) built into most modern ATA and SCSI
hard disks. It is derived from the smartsuite package, and includes support
for ATA/ATAPI-5 disks. It should run on any modern Linux system.
Your init is not systemd it seems:
what do below ones show?
readlink -f /sbin/init
/sbin/init --version
apt policy systemd
dpkg -S bin/init
Oh I realised later that above syslogd could be part of that busybox-syslogd package.
Then klogd cant be removed/disabled ... I think.
You can list its content with below:
dpkg -L busybox-syslogd
Or search to which package it belongs:
dpkg -S bin/syslogd
Am not that familiar yet with busybox
Only one that I know of is my router:
$ busybox
BusyBox v1.17.4 (2020-06-09 09:41:02 CST) multi-call binary.
Copyright (C) 1998-2009 Erik Andersen, Rob Landley, Denys Vlasenko
and others. Licensed under GPLv2.
See source distribution for full notice.
Usage: busybox [function] [arguments]...
or: function [arguments]...
BusyBox is a multi-call binary that combines many common Unix
utilities into a single executable. Most people will create a
link to busybox for each function they wish to use and BusyBox
will act like whatever it was invoked as.
Currently defined functions:
[, [[, arp, ash, awk, basename, blkid, cat, chmod, chown, chpasswd,
clear, cmp, cp, crond, cut, date, dd, devmem, df, dirname, dmesg, du,
e2fsck, echo, egrep, env, ether-wake, expr, fdisk, fgrep, find, flock,
free, fsck, fsck.ext2, fsck.ext3, fsck.minix, fsync, grep, gunzip,
gzip, head, ifconfig, insmod, ionice, kill, killall, klogd, less, ln,
logger, login, ls, lsmod, lsusb, md5sum, mdev, mkdir, mke2fs,
mkfs.ext2, mkfs.ext3, mknod, mkswap, modprobe, more, mount, mv,
netstat, nice, nohup, nslookup, pidof, ping, ping6, printf, ps, pwd,
readlink, renice, rm, rmdir, rmmod, route, sed, setconsole, sh, sleep,
sort, strings, swapoff, swapon, sync, syslogd, tail, tar, telnetd,
test, top, touch, tr, traceroute, traceroute6, true, tune2fs, udhcpc,
umount, uname, unzip, uptime, usleep, vconfig, vi, watch, wc, which,
zcat, zcip
$ echo $SHELL
/bin/sh
$ readlink -f /bin/sh
/bin/busybox