tracker-store uses 100% cpu and fills all the disk space in home partition

[EcLiPsE]

I recently noticed that several tracker related programs were using 100% cpu on my system and in particular tracker-store filled up completely all the available disk space in my home partition.

I rerun tracker-preferences to make sure that indexing would not proceed if less than 20% of free space is available, but it seems that tracker-store ignores it. After I noticed that I had left only 44Kb free on the home partition, I deleted the cache from another program and I noticed that tracker-store filled up again the disk space that I freed few minutes before.

This is the content of my .cache/tracker/ directory:
totale 772068
-rw-r--r-- 1 eclipse eclipse 3072 2011-06-26 22:56 contents.db
-rw-r--r-- 1 eclipse eclipse 11 2012-01-04 21:24 db-locale.txt
-rw-r--r-- 1 eclipse eclipse 2 2011-12-12 04:02 db-version.txt
-rw-r--r-- 1 eclipse eclipse 3072 2010-12-05 21:29 email-contents.db
-rw-r--r-- 1 eclipse eclipse 3072 2009-05-06 19:16 email-fulltext.db
-rw-r--r-- 1 eclipse eclipse 2359292 2011-06-26 22:54 email-index.db
-rw-r--r-- 1 eclipse eclipse 34816 2010-12-05 21:29 email-meta.db
-rw-r--r-- 1 eclipse eclipse 17062912 2011-06-26 19:30 file-contents.db
-rw-r--r-- 1 eclipse eclipse 15014912 2010-01-12 20:26 file-fulltext.db
-rw-r--r-- 1 eclipse eclipse 75609207 2011-06-26 22:54 file-index.db
-rw-r--r-- 1 eclipse eclipse 37027840 2011-06-26 19:31 file-meta.db
-rw------- 1 eclipse eclipse 1180048 2009-05-05 22:57 file-update-index.db
-rw-r--r-- 1 eclipse eclipse 28376064 2011-10-20 09:34 fulltext.db
-rw-r--r-- 1 eclipse eclipse 1024 2012-01-04 21:24 meta.db
-rw-r--r-- 1 eclipse eclipse 4653056 2012-01-04 21:26 meta.db-shm
-rw-r--r-- 1 eclipse eclipse 608064304 2012-01-04 21:26 meta.db-wal
-rw-r--r-- 1 eclipse eclipse 11 2011-12-12 04:02 miner-applications-locale.txt
-rw-r--r-- 1 eclipse eclipse 345110 2012-01-04 21:24 ontologies.gvdb

and this is how much space I have left on my home partition after I KILLED the tracker-store process:
/dev/sda8 31G 30G 418M 99% /home
---
ApportVersion: 1.23-0ubuntu4
Architecture: i386
DistroRelease: Ubuntu 11.10
Package: tracker 0.10.24-1build2
PackageArchitecture: i386
ProcEnviron:
 LANGUAGE=it_IT:it:en_GB:en
 PATH=(custom, user)
 LANG=it_IT.UTF-8
 SHELL=/bin/bash
ProcVersionSignature: Ubuntu 3.0.0-14.23-generic 3.0.9
Tags: oneiric
Uname: Linux 3.0.0-14-generic i686
UpgradeStatus: Upgraded to oneiric on 2011-10-20 (76 days ago)
UserGroups: adm admin audio cdrom dip fuse lpadmin plugdev polkituser pulse-access sambashare src video

[EcLiPsE (softinn72)]

apport information

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.

[Paul (reclusivegeek)]

I have the exact same problem.

I have temporarily fixed the problem by stopping tracker-store starting by editing /etc/xdg/autostart/tracker-store.desktop and changing the line X-GNOME-Autostart-enabled=true to X-GNOME-Autostart-enabled=false and making the same changes to /etc/xdg/autostart/tracker-miner-fs.desktop

I guess I am just going to have to live without desktop search until this bug gets fixed.

[Markcortbass (markcortbass)]

I confirm this serious bug with Ubuntu 14.04 LTS with the latest updates.

[Martyn Russell (martyn-lanedo)]

Really, you're using Tracker 0.10?

You have / had databases in the cache dir which haven't been used by Tracker in about half a decade, probably longer.

I would suggest you upgrade and update your config (at least to the default) to avoid these sort of problems as best as possble.

[Dee (dmusil-x)]

As recommended workaround i get this:
gsettings set org.freedesktop.Tracker.Miner.Files crawling-interval -2
gsettings set org.freedesktop.Tracker.Miner.Files enable-monitors false
tracker-control -r
This will disable the constant indexing and cleanup the indexers database.

[CRAFT (craft37)]

It affects me. tracker-miner-fs costs me 23% of CPU and 296mb without any desktop app opened.
tracker-extract eat pretty same amount.

CPU Intel® Core™ i7 CPU M 640 @ 2.80GHz × 4
Ubuntu 16.04

[CRAFT (craft37)]

Thanks Dee, it works for me.

[Luca (zapduke)]

Never had this problem before but now it's happening constantly on Ubuntu 16.04, I had to shut down the processes.

[Harvey Houston (ubuntu-who)]

I have seen where this stays on top of the list when running `top` in a terminal. As a temporary workaround, use this link to solve your problems. Run each line of the script box in a terminal:
https://askubuntu.com/questions/346211/tracker-store-and-tracker-miner-fs-eating-up-my-cpu-on-every-startup#348692

[Luis Andrés Gonzalez (luisandresgonzalez)]

In my case, as a workaround execting

cpulimit -l 20 -p <process id>

Helped in reducing the process CPU usage to 20%.

Please, let me know if this is an unefficient way to do it.

[Dan (dannyhajj)]

I'm not sure if this is the same issue or if it should be reported on its own. But in my case, the miner was filling my / partition by filling the syslog with repeated logs.

The following message gets repeated endlessly:

Jan 25 12:44:04 localhost tracker-miner-f[14902]: File 'file:///path/to/directory/in/Downloads' has been reenqueued more than 2 times. It will not be indexed.
Jan 25 12:44:04 localhost tracker-miner-f[14902]: File 'file:///path/to/directory/in/Downloads/file' has been reenqueued more than 2 times. It will not be indexed.
Jan 25 12:44:04 localhost tracker-miner-f[14902]: (Sparql buffer) Error in task 0 (file:///home/danny/Downloads) of the array-update: UNIQUE constraint failed: nie:DataObject.nie:url
Jan 25 12:44:04 localhost tracker-miner-f[14902]: Could not execute sparql: UNIQUE constraint failed: nie:DataObject.nie:url
Jan 25 12:44:04 localhost tracker-miner-f[14902]: (Sparql buffer) Error in task 0 (file:///home/danny/Downloads) of the array-update: UNIQUE constraint failed: nie:DataObject.nie:url
Jan 25 12:44:04 localhost tracker-miner-f[14902]: Could not execute sparql: UNIQUE constraint failed: nie:DataObject.nie:url
Jan 25 12:44:04 localhost tracker-miner-f[14902]: (Sparql buffer) Error in task 0 (file:///home/danny/Downloads) of the array-update: UNIQUE constraint failed: nie:DataObject.nie:url
Jan 25 12:44:04 localhost tracker-miner-f[14902]: Could not execute sparql: UNIQUE constraint failed: nie:DataObject.nie:url

It's currently 12:59, the /var/log/syslog was created at 12:44 and it's already 1.4G.

$ ls -lhS /var/log/sys*
-rw-r----- 1 syslog adm 5.8G Jan 25 10:33 /var/log/syslog.1
-rw-r----- 1 syslog adm 1.4G Jan 25 12:44 /var/log/syslog
-rw-r----- 1 syslog adm 42M Jan 23 09:51 /var/log/syslog.3.gz
-rw-r----- 1 syslog adm 40M Jan 24 09:46 /var/log/syslog.2.gz
-rw-r----- 1 syslog adm 34M Jan 17 09:53 /var/log/syslog.7.gz
-rw-r----- 1 syslog adm 26M Jan 18 10:03 /var/log/syslog.6.gz
-rw-r----- 1 syslog adm 22M Jan 22 10:16 /var/log/syslog.4.gz
-rw-r----- 1 syslog adm 7.3M Jan 19 09:56 /var/log/syslog.5.gz

The message appears over 1.6 million times in the syslog file, and over 7.3 million times in the syslog.1 file:
$ grep "has been reenqueued more than 2 times." /var/log/syslog | wc -l
1681093
$ grep "has been reenqueued more than 2 times." /var/log/syslog.1 | wc -l
7306358

I had only noticed it when my system started reporting that it is low on disk space.

$ tracker --version
Tracker 2.0.1

This program is free software and comes without any warranty.
It is licensed under version 2 or later of the General Public License which can be viewed at:

  http://www.gnu.org/licenses/gpl.txt

$ uname -a
Linux danny-laptop-91 4.13.0-31-generic #34-Ubuntu SMP Fri Jan 19 16:34:46 UTC 2018 x86_64 x86_64 x86_64 GNU/Linux

$ lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 17.10
Release: 17.10
Codename: artful

I'm

[Dan (dannyhajj)]

I have accidentally removed part of the last line, but I wanted to mention that I'm using the default Ubuntu Gnome installation.

[Pedro Guimarães (pguimaraes)]

Using focal desktop and I am seeing the same issue.
I was seeing tracker-store and tracker-miner-fs taking an entire core of my notebook.
Tried: tracker reset --hard and reboot.
Did not work for me.

Only effective solution was to disable search engine on "Settings".

[teo1978 (teo8976)]

I'm having this issue. I googled for the name of the process that was eating up 100% of CPU and, surprise! A bug reported in 20-fucking-12 and still not fixed!

Great job, Ubuntu!

[michael kosko (punk80)]

This issue is still present. I just did a fresh 20.04.4 installation and tracker still used high CPU at system startup until I disabled it.