Turris OS 3.11 is out!

[miska]

Dear Turris users,

we just released Turris OS 3.11. This release is pretty big and brings quite some new or redesigned features, but it spend quite some time in RC, so we hope that all major bugs were found and mercilessly squashed.

One of the cool features in celebration of the nice round release number is experimental support for samba 4. You can try to replace samba36 with it.

Our web interface Foris was migrated to Python3 and we are switching to a new approach to the first time configuration of the router. Regarding the configuration itself, some parts were rewritten a little bit and we are also testing different approach to displaying results in Pakon plugin. We tried to integrate various third party web applications better - you can check it out on base url of your router (for example http://192.168.1.1/).

Apart from various updates, we are also shipping first part of our new data collecting and firewall distributing system Sentinel in this release. How to test it and more information in general about it can be found in separate post.

Full release notes are as follows:

• foris: migration to python3
• foris: wizard replaced by guide
• foris: small cleanups and rearranging
• sentinel: first parts of new data collection engine
• samba4: experimental support
• better integration of third party web applications (details: Webapps - documentation)
• various package updates

Hope you will enjoy the new features and as always, we welcome any feedback regarding this release.

[anon50890781]

INFO:Unpacking download packages
INFO:Checking for file collisions between packages
line not found
line not found
line not found
line not found
line not found
line not found
DIE:
[string “transaction”]:323: [string “transaction”]:149: Collisions:
• /sbin/ntpd: busybox (new-file), ntpd (existing-file)
Aborted

How to solve that?

[Pepe]

You don’t need to have package ntpd, which you have installed as it is provided by busybox.
What you need to do is to remove package ntpd, which you have installed and run pkgupdate again.

Also, it is not possible to install it. See following error:

Collected errors:
 * check_data_file_clashes: Package ntpd wants to install file /sbin/ntpd
	But that file is already provided by package  * busybox
 * opkg_install_cmd: Cannot install package ntpd.

[protree]

Full Version of netcat (Installed through package netcat from opkg) was overwritten/deleted during update. Had to manually delete package netcat (Printed error that /usr/bin/nc couldn’t be deleted because its missing) and install it again. Now it is working.

[sammy_cda]

The router status page indicates there are no Active UPnP Redirects while Transmission has no problem working through the Windows firewall.

I’m not sure why the status page no longer indicates there are no active UPnPs anymore?

[tristone]

Since the update I have repeated error message related to BTRFS. This started to happen with the update, I checked in /var/log/messages

2018-12-10 18:55:24 err netdata[8046]: BTRFS: failed to read '/sys/fs/btrfs/0df3df80-9c53-438e-9905-07e5de714a81/allocation/data/total_bytes'
2018-12-10 18:56:01 info /usr/sbin/cron[13656]: (root) CMD (/usr/bin/rainbow_button_sync.sh)
2018-12-10 18:56:01 info /usr/sbin/cron[13657]: (root) CMD (nethist_stats.lua)
2018-12-10 18:56:24 err netdata[8046]: BTRFS: failed to read '/sys/fs/btrfs/0df3df80-9c53-438e-9905-07e5de714a81/allocation/data/total_bytes'
2018-12-10 18:57:01 info /usr/sbin/cron[13715]: (root) CMD (/usr/bin/rainbow_button_sync.sh)
2018-12-10 18:57:26 err netdata[8046]: BTRFS: failed to read '/sys/fs/btrfs/0df3df80-9c53-438e-9905-07e5de714a81/allocation/data/total_bytes'
2018-12-10 18:58:01 info /usr/sbin/cron[13750]: (root) CMD (nethist_stats.lua)
2018-12-10 18:58:01 info /usr/sbin/cron[13751]: (root) CMD (/usr/bin/rainbow_button_sync.sh)
2018-12-10 18:58:26 err netdata[8046]: BTRFS: failed to read '/sys/fs/btrfs/0df3df80-9c53-438e-9905-07e5de714a81/allocation/data/total_bytes'

[madko]

Hi,

I’ve just received this notification from my router:

##### Error notifications #####
Updater selhal: 

[string "transaction"]:323: [string "transaction"]:149: Collisions:

• /sbin/ntpd: busybox (new-file), ntpd (existing-file)

##### Update notifications #####
 • Installed version 20180409-2 of package ca-bundle
 • Installed version 7.62.0-2 of package libcurl
 • Installed version 60.4.6-3.6-2 of package updater-ng

I will try to remove the ntpd package since it seems that ntp daemon is now provided by busybox.

Best regards,
Edouard

[sammy_cda]

Regarding my UPnP post above:

I was able to resolve this. It turns out that one of the three packages, minidlna, miniupnpc or miniupnpd, I don’t remember which, wasn’t installed after the update.

I installed it and now UPnP seems to be working well again.

[ge-man]

I’m having a issue with the update never completing.

The last output I see in /var/log/message from the updater is

2018-12-10 20:39:56 debug updater[5829]: backend.lua:1021 (script_run):
2018-12-10 20:39:56 debug updater[5829]: backend.lua:1009 (script_run): Running postinst of kmod-tun
2018-12-10 20:39:56 debug updater[5829]: src/lib/interpreter.c:320 (lua_run_generic): Command: /usr/lib/opkg/info//kmod-tun.postinst configure

after that nothing happens (I’ve waited ~20 minutes). It never seems to complete. I am not sure at what state my updater is in, I rebooted around the same time the new update was released. the cpu is mostly idle and no io activity is ongoing as far as I can tell from top

Edit: Ok after some digging I figured it was maybe not kmod-tun which hanged. If I sent a SIGINT to the opkg-trans and ran it manually it always got stuck on /usr/lib/opkg/info//ddns-scripts.postinst. So I removed that with opkg install ddns-scripts luci-app-ddns and re-ran the update successfully. After that I had to force reinstall them for the updater to stop throwing warnings with opkg --force-reinstall opkg ddns-scripts luci-app-ddns

Now I am unable to login as root with ssh instead.

Edit2:

So the root login was because I used zsh as shell for root. Before it was at /usr/zsh now it is at /usr/bin/zsh. Easily fixed as I had a user account I could login with and su -s /usr/bin/zsh and then update it in /etc/passwd

Anyways, even with the issues. It’s great to see the turris getting updates, it was why I bought it. If I wanted an outdated router which never “breaks” I would have gotten something else. I see people being upset, and I can understand that is frustrating. But upgrading customised openwrt routers is probably a NP problem and if one prefers to wait a bit they provide both the ability to disable updates, to approve updates or to do it later with a delay. Keep up the good work!

[DIKKEHENK]

Ok, installed 3.11. Help me out here…Since this nOOb does not understand the data collection part.
Should i activate, or de activate data collection?
Should i have Data Collection Software for participation in data collection and dynamic distributed firewall. marked?

pakon runs fine now! Thank you :). not sure about this sentinel part… SSH honey is on.,

thx.

Best, Dikke

[DIKKEHENK]

yep, i have the same.

[Mordorf]

Yet again another botched update. This really is getting beyond a joke. After pretty much every update I have to go and fix stuff to make it all work again.
Return of Update broke my router this time around.
Also missing miniupnpc package so as mentioned by someone else had to manually install to get upnp working.

Seriously thinking of selling this piece of junk. Anyone interested in a non wifi model going cheap pm me.

[Pepe]

Removing ntpd package should work for you.

[Mordorf]

Why remove ntpd?
Doesn’t seem to be installed for me anyway.

[madko]

removing ntp package resolved the issue indeed.

I also got this one:

INFO:Running post-install and post-rm scripts
Output from resolver-conf.postinst:
chown: unknown user/group kresd:kresd
uci: Entry not found

Is it ok ?

and this:

+ /etc/init.d/lighttpd restart
Output from pakon.postinst:
Traceback (most recent call last):
  File "/usr/libexec/pakon-light/create_db.py", line 24, in <module>
    c.execute('CREATE INDEX IF NOT EXISTS archive1 ON traffic(src_mac, start, COALESCE(app_hostname,dest_ip))')
sqlite3.OperationalError: database or disk is full
Error: database or disk is full
Error: database or disk is full
Command failed: Not found

with:

# df -h
Filesystem                Size      Used Available Use% Mounted on
/dev/mmcblk0p1            7.3G      2.4G      4.9G  33% /
tmpfs                  1010.9M    250.1M    760.7M  25% /tmp
tmpfs                   512.0K         0    512.0K   0% /dev

I’m afraid to reboot. Is it safe ?

[jklaas]

Well, my Statistics page was mangled again after the update after quite a few updates where it wasn’t. My self-installed graph units were blown away (/usr/lib/lua/luci/controller/luci_statistics/luci_statistics.lua and /usr/lib/lua/luci/controller/luci_statistics/luci_statistics.lua were overwritten) but also there seemed to be a regression of /usr/lib/lua/luci/statistics/rrdtool/definitions/df.lua

I keep forgetting which file needs to be removed to update LuCI after the files are restored/updated. It’s /tmp/luci-indexcache

I also encountered a scary error message on the first refresh after reboot that went away after I reloaded the page. I captured a PDF of the page, but I’m not sure it’s worth posting since it went away right away.

I think a lot of folks have YaMON installed. How hard would it be to have a large “button” at the landing page go there along with the “Foris” and “LuCI” buttons? Is there any way to keep it from automatically going to the Foris page after 5 seconds?

Pakon filtering seems to work again, though I’m not entirely sold on the Overview. Pretty, but I’m not sure how really useful that is.

[Pepe]

That’s related to netdata, because it can’t read the value. You can disable in it.

The best way, how you can do it is download netdata.conf, which you can download from here http://192.168.1.1:19999/netdata.conf and replace it in your router /etc/netdata/netdata.conf

In the replaced file, you need to uncomment and change the value of it.

There you can see:

 	# /sys/fs/btrfs = yes

Uncomment it by just deleting the # and instead of yes, write there no.

So, in the end, it will look like this:

	 /sys/fs/btrfs = no

Restart netdata and errors from the netdata are gone.

[Pepe]

chown: unknown user/group kresd:kresd
uci: Entry not found

Thanks for reporting, @paja will look into it! About the second issue will @mpetracek tell you more.