too many strange issues and missing packages. i reverted back to 3.7.2. guess i’ll wait and see how 3.8.1 plays out. getting tired of these .x releases essentially being betas.
horrible update!
one of my omnia stock 3.7.2 no custom changes… unvaible after 3.8 update… I must revert it back… SSH not accesibile, luci and foris also… wifi not avaible
Old update error back.
Upon updating to OS 3.8, I started getting these errors again when attempting to fetch the daily update…
Error from 2017/09/15 21:55:03
Updater failed:
inconsistent: Package luci-ssl requires package libustream-polarssl that is not available.
I would appreciate any suggest this issue.
Thanks!
Robert
btw. my details are as follows:
Device Turris Omnia - RTROM01
Turris OS version 3.8
Kernel version 4.4.87-cb5e816fa6b1a6b5342df69755869d71-2
@rguerra, this was addressed already:
Just remove luci-ssl and eventually install lighttpd-https-cert.
Previous stable version was 3.7.5 so you had to have disabled updater (ignoring huge warning) so here is your custom setup. But I kind of missed what is your problem with this update.
We know that some wifi cards are problematic with TO so if you are experiencing problems with 5Ghz please contact our support. There are two solutions, you can send us TO and we will replace some capacitors on board to different values or you can move 5Ghz wifi to left most pcie slot but for that you need longer pigtail which we can potentially provide. So contact our support.
I am sorry for not informing about that. It was dropped as part of the userlists cleanup. It had no out of the box use on TO and was just artefact of Turris 1.x. If you are using it then you have to install it by hand and if you are at it remove kresd if you are not using it.
Since update is problem with both wi-fi. with 5Ghz and 2.4Ghz. Until update there wasn’t any problem with any card
I get the following error when I try to change scheduled tasks in Luci:
# DO NOT EDIT THIS FILE - edit the master and reinstall.
# (/etc/crontabs/root installed on Sat Sep 16 11:05:38 2017)
# (Cron version V5.0 -- $Id: crontab.c,v 1.12 2004/01/23 18:56:42 vixie Exp $)
Typical Linux distributions have a notion of orphaned packages: packages that where installed as dependency but no longer needed. They are not uninstalled but you may list them to remove them by hand. I think it would be a cleaner behavior.
Anyway thanks for the explanation.
I’m getting the same problem, has support for transmission-web been intentionally removed. Problems like this defeat the purpose of a router which has automatic security updates.
No it wasn’t. It’s repeated failure of OpenWRT build system. Will be fixed in 3.8.1.
We have that too but instead of just waiting and letting you to maintain and remove it on your own we just remove it immediatly. That is because partially we manage your router remotely and we need this feature. If you want to check if updatr won’t remove some package then use updater approvals instead.
Which capacitors need replacing? I can do this myself.
Thanks, it was in auto.lua, so apparently I had to install it at some point. It would be great to have such things mentioned in the release notes.
Anyway this made me run the updater from the command line and I got another error:
ERROR:Failed operations:
luci-app-sqm/postrm:
foris-openvpn-plugin-l10n-cs/postrm: + [ -n ]
The error messages doesn’t seem to be much helpful…
Speaking of that there is quite a lot of things what seem to be unhandled errors in the updater output:
Output from foris-diagnostics-plugin-l10n-cs.postinst:
- [ -n ]
- /etc/init.d/lighttpd restart
grep: /proc/16556/cmdline: No such file or directory
grep: /proc/16558/cmdline: No such file or directory
grep: /proc/16574/cmdline: No such file or directory
grep: /proc/16578/cmdline: No such file or directory
grep: /proc/16579/cmdline: No such file or directory
grep: /proc/16580/cmdline: No such file or directory
Output from sqm-scripts.postinst: - [ -n ]
- /etc/init.d/sqm-scripts enable
/usr/lib/opkg/info//sqm-scripts.postinst: /usr/lib/opkg/info/sqm-scripts.postinst-pkg: line 1: /etc/init.d/sqm-scripts: not found
SQM: Stopping SQM on eth1
SQM: Starting SQM script: simple.qos on eth1, in: 150000 Kbps, out: 150000 Kbps
SQM: simple.qos was started on eth1 successfully
Output from knot-resolver.postinst: - [ -z ]
- /etc/init.d/dnsmasq restart
udhcpc: started, v1.25.1
udhcpc: sending discover
udhcpc: no lease, failing
udhcpc: started, v1.25.1
udhcpc: sending discover
udhcpc: no lease, failing - sleep 2
- /etc/kresd/kresd.postinst.sh
- uci -q get resolver.kresd.static_domains
- stat_dom=
- [ ! -z ]
- /etc/init.d/resolver restart
/etc/rc.common: eval: line 1: get_dnsmasq_dhcp_script: not found - rm /etc/kresd/kresd.postinst.sh
- /etc/kresd/convert_config.sh
uci: Entry not found - rm /etc/kresd/convert_config.sh
Output from resolver-conf.postinst:
/etc/rc.common: eval: line 1: get_dnsmasq_dhcp_script: not found
Output from luci-app-statistics.postinst:
configfile: stat (/etc/collectd/conf.d) failed: No such file or directory
rrdtool plugin: RRASingle = true: creating only AVERAGE RRAs
Output from pakon-guts.prerm:
Command failed: Not found
Output from foris-openvpn-plugin-l10n-cs.postrm: - [ -n ]
Output from unbound.postrm:
/etc/rc.common: eval: line 1: get_dnsmasq_dhcp_script: not found
I have the same issues as some mentioned here before. Using Turris 1.x. After update
- cannot access web GUI. Found out it is trying to use same port as OpenVPN (443).
- 5G WiFi stopped working although it acts as enabled.
- router reboots in few hours intervals. Approximately once per 3 hours.
Any tips how to solve this?
Must say that this update is a tragedy… 
Thanks for finding out - I do have the very same configuration! As I didn’t find out wich config-file is responsible for that behaviour I opened another thread -> Turris OS v.3.8 - changed ports for lighttpd
@ turris developers: please consider for future updates customers NOT using standard ports! Reasons for that are various and obvious.
I strongly recommend for future updates to tell us users what you exactly changed (have a look at Discussion: updater-configuration)
@ladislav.balik by pure curiousity - how did you find out?
edit: solved that problem. For others with that same problem have a look into the linked thread.
The other problems don’t occur for me - wifi is working without problems, router is up for 34 hours by now (and last time I rebooted manually).
This Update seems to crash my Turris Omnia completely!
I was busy a couple days and when I came back home yesterday I had no network connection in my house. I wasn’t able to even ping the Omnia, so I decided to recover with a previous settings-backup. Everything worked back normal.
Then today I received again the mail notification, that the Omnia was updated, but I need to restart the router to take effect of the update. At this point, my Omnia was already crashed. Same behavior - no internet connection, no ping, no dhcp, etc. There is even no link on the port (switch has no link to the router).
The Omnia LED are first normal (some are white blinking), then they turn to completely blue (all at the same time), then just power-LED and completely blue again.
Because I’m not able to connect to the Turris Omnia in this state, I’m not even know whats exactly the problem.
Do you have any ideas?