Network overview page failed

Tried to do following, but did not solve issue:

sudo service pihole-FTL stop
sudo mv /etc/pihole/pihole-FTL.db /etc/pihole/pihole-FTL-old.db
sudo service pihole-FTL start 

IP6 disabled on this box.

Expected Behaviour:

PI-HOLE WUI section Tools/Network produced following error:

An unknown error occurred while loading the data.
{"error":{"key":"not_found","message":"Not found","hint":"/api/network/devices"},"took":0.00033926963806152344}

  • Linux 13 (trixie) / 6.18.34+rpt-rpi-v8
  • Rspberry Pi 4

Actual Behaviour:

2026-07-24 03:08:00.089 ERROR SQLite3: database corruption at line 76524 of [c88b22011a] (11)
2026-07-24 03:08:00.089 ERROR SQLite3: database disk image is malformed in "SELECT vendor FROM macvendor WHERE mac IN (substr(upper(?1),1,8),substr(upper(?1),1,9)||substr(upper(?1),10,1)||'0/28',substr(upper(?1),1,12)||substr(upper(?1),13,1)||'0/36') ORDER BY length(mac) DESC LIMIT 1;"...
2026-07-24 03:08:00.089 ERROR getMACVendor("1e:39:5d:3e:19:4c") - SQL error prepare "SELECT vendor FROM macvendor WHERE mac IN (substr(upper(?1),1,8),substr(upper(?1),1,9)||substr(upper(?1),10,1)||'0/28',substr(upper(?1),1,12)||substr(upper(?1),13,1)||'0/36') ORDER BY length(mac) DESC ...
2026-07-24 03:08:00.089 ERROR getMACVendor("1e:39:5d:3e:19:4c") - SQL error prepare "SELECT vendor FROM macvendor WHERE mac IN (substr(upper(?1),1,8),substr(upper(?1),1,9)||substr(upper(?1),10,1)||'0/28',substr(upper(?1),1,12)||substr(upper(?1),13,1)||'0/36') ORDER BY length(mac) DESC ...
2026-07-24 03:08:00.089 WARNING Database /etc/pihole/pihole-FTL.db is damaged and cannot be used.
2026-07-24 03:08:04.194 WARNING Failed to open database in networkTable_readDevices()
2026-07-24 03:08:18.285 ERROR format_messages() - Database not available
2026-07-24 03:08:18.285 WARNING API: An internal error occurred while formatting the messages array. (key: internal_error)

Debug Token:

https://tricorder.pi-hole.net/k54ntO0J/

The first 3 lines are saying that macvendor.db database file is corrupted:

Try sudo pihole -r to repair the installation. This should replace the corrupted file.


Then, another log line says pihole-FTL.db is damaged.

Your commands to probably already fixed pihole-FTL.db database.

If the issue remains even after repairing pihole-FTL.db database and reinstalling other Pi-hole files, you will need to check why there are so many corrupted files on your system. I suspect your SD card is failing.

SD card is first what replaced. Issue persist with new one. So I absolutely lost why SQLdb still failing. Sometimes log entry mentioned tmp partition - to small. New SD size is 128Gb.
Is it possible I have faulty RP4??

tried sudo pihole -r

immediately get following in the log:
2026-07-25 04:21:00.014 ERROR SQLite3: database corruption at line 76524 of [c88b22011a] (11)

2026-07-25 04:21:00.014 ERROR SQLite3: database disk image is malformed in "SELECT vendor FROM macvendor WHERE mac IN (substr(upper(?1),1,8),substr(upper(?1),1,9)||substr(upper(?1),10,1)||'0/28',substr(upper(?1),1,12)||substr(upper(?1),13,1)||'0/36') ORDER BY length(mac) DESC LIMIT 1;"...

2026-07-25 04:21:00.014 ERROR getMACVendor("ac:0b:fb:1e:12:18") - SQL error prepare "SELECT vendor FROM macvendor WHERE mac IN (substr(upper(?1),1,8),substr(upper(?1),1,9)||substr(upper(?1),10,1)||'0/28',substr(upper(?1),1,12)||substr(upper(?1),13,1)||'0/36') ORDER BY length(mac) DESC ...

2026-07-25 04:21:00.014 WARNING Database /etc/pihole/pihole-FTL.db is damaged and cannot be used.

Will replace SD card and do clean install

The Raspberry Pi OS Developers have made some default changes about both the /tmp/ and /var/log/ stuff over the last couple of years :

  • /tmp/ got moved to a file I believe ?!
  • /var/log/ got moved to RAM instead of the local storage.

And I am guessing both of these can be changed via raspiconfig or whatever that command was...

Where did you read that?

Has anything changed with Trixie compared to below Bookworm?

$ hostnamectl
[..]
Operating System: Raspbian GNU/Linux 12 (bookworm)
$ findmnt --target /tmp
TARGET SOURCE         FSTYPE OPTIONS
/      /dev/mmcblk0p2 ext4   rw,noatime
$ findmnt --target /var/log
TARGET SOURCE         FSTYPE OPTIONS
/      /dev/mmcblk0p2 ext4   rw,noatime
$ lsblk /dev/mmcblk0p2
NAME      MAJ:MIN RM  SIZE RO TYPE MOUNTPOINTS
mmcblk0p2 179:2    0 14.3G  0 part /
$ file /dev/mmcblk0p2 /tmp /var/log
/dev/mmcblk0p2: block special (179/2)
/tmp:           sticky, directory
/var/log:       directory

Not here:

rdweb@pi4b:~ $ hostnamectl | grep Operating
Operating System: Debian GNU/Linux 13 (trixie)

rdweb@pi4b:~ $ findmnt --target /tmp
TARGET SOURCE FSTYPE OPTIONS
/tmp   tmpfs  tmpfs  rw,nosuid,nodev,size=4003732k,nr_inodes=1048576

rdweb@pi4b:~ $ findmnt --target /var/log
TARGET SOURCE         FSTYPE OPTIONS
/      /dev/mmcblk0p2 ext4   rw,noatime

rdweb@pi4b:~ $ file /dev/mmcblk0p2 /tmp /var/log
/dev/mmcblk0p2: block special (179/2)
/tmp:           sticky, directory
/var/log:       directory

A noticeable difference is that above /tmp directory got its own mount in RAM now compared to Bookworm.

$ man tmpfs
[..]
DESCRIPTION
       The  tmpfs  facility  allows the creation of filesystems whose con‐
       tents reside in virtual memory.  Since the files on  such  filesys‐
       tems typically reside in RAM, file access is extremely fast.

In the past at several forums, maybe some of it even here on the forum :slight_smile:

Your reaction suggests I am wrong and in case I am then maybe I have mixed up Debian and Ubuntu specific changes with Raspberry Pi OS changes ?!

Stuff happens... I am not as sharp as I used to be sadly due to age and other reasons... :wink:

I can only see the /tmp folder moved to a RAM mount.
The /var/log folder still resides on the root "/" mount (on disk/SD card):

Maybe you're mistaking with the systemd journals that got moved to RAM a couple of major releases ago?

Cant say I agree with those decisions.
Older less beefy Pi models will struggle with available RAM.

Maybe I am confused about that one too...?!