Turris OS 6.0 is released

Re: 6.0 and SDIO.
I have sent a support ticket with diagnostics (1498385).

2 Likes

Started testing it, two points:

  1. eth0 interface is now not used at all. As I understand, it will stay this way?
  2. /bin/tar is not a symbolic link to busybox now, but to /usr/libexec/tar-gnu. In your script /usr/bin/maintain-config-backup you use /usr/bin/tar as a command, but this link is now missing completely.

Kind regards,

Finally got around to confirm/initiate the update, took a while (expected for replacing basically everything) and after an additional? manual reboot that reforis informed was required everything seems to work.
Sure this is a rather pedestrian configuration, but sqm, WiFi, pakon, collectd, ssh access, everything just seems to work. To my joy IPv6 also works out of the box so far (dual stack via PPPoE, with DHCPv6 in use for distributing IPv6 prefixes). OpenVPN seems to work (but I only tested from inside my LAN).
So far so good.
Netmetr finally reports RTT values on my DualStack link, but I guess with netmtr not being the official cz speedtest this might or might not be a long term solution. (It is not that I need automated speedtests, but they are quite intersting).
ATLAS also works, I can successfully recruit my omnia as probe in IPv4 and IPv6 measurements.

Nice touch, clicking the “turris” text on the top left in LuCI will get me back to the selection screen, is that new? Or I might just be looking closer now and just discovered an old feature.

The green check persists.

Hi @MichalCirrus,

You are right, thanks for noticing this.

I have already filled in an issue and fix will hopefully be available soon.

1 Like

There is issue for that:

It will not stay like this, but the situation with MultiCPU DSA was quite complicated and there were several attempts to have it upstream to Linux kernel, but situation improved and there were some parts merged. What we need right now is to let’s say rework our multiDSA support which we had and make it compatible with upstream version. I am afraid that it requires even newer kernel that what we are having now.

3 Likes

Hello,

I’m trying to update again from 5.4.4 to 6.0.1.
I need to apply this solution (thanks @robinb ) for start lighttpd correctly :

turris procd: Instance lighttpd::instance1 s in a crash loop 6 crashes, 0 seconds since last crash

Apparently in the 30-rrdtool.conf it was crashing of missing the home_dir:
`## Path to the rrdtool database. You can override this in conditionals.’
‘##’
rrdtool.db-name = home_dir + “/lighttpd.rrd”’

Resolved by ## out the line rrd.db-name. Manually started lighttpd and things seem to be working again.

Can you consider this issue ?

Kind regards.

Same here,

No SDA the first day.
No LAN the next day.

The status LED is ON all the time instead of hearth beet.

Just would like to add my upgrade experience:

  1. Waited with upgrade couple of hours (everybody can select approval for updates).
  2. After install, turris had to reboot 2 times.
  3. Everything works

Thank you @Pepe / @miska and whole Turris team!

2 Likes

Just to say the same (had to reboot twice)
Everything seems OK, including my working wireguard tunnel to another Turris omnia instance (not yet migrated)

Thanks a lot to the Turris Team … now time to mess the working config trying to set up VLANs :slight_smile:

1 Like

I can confirm the same, after a restart of the fw-logs it is green for a short time but then red again.

hello i try to upgrade from TurrisOS 5.4.4 to 6.0.1 on turris omnia 2020 but it’s imposible
i removed all not recognized packages:

kmod-ipt-nathelper-rtsp
iptgeoip
kmod-ipt-compat-xtables
libuvc
asterisk16
Opencv

i disabeled IPV6 as my mobile isp does not support it.
I tried dnssec enableled disabeled forwarded or not but i alway have a error :

DIE:
download: Download of https://repo.turris.cz/hbs/omnia/packages/XXX.ipk failed: Couldn’t resolve host ‘repo.turris.cz’
Aborted

thank you for your help

The LuCI “dashboard” doesn’t look good anymore. At least this update didn’t seem break anything else for me.

I have got exactly same error on Turris 1.1.

Can someone help me?

There is output from pkgupdate -e DBG :

At the begining it looks that resolving of https://repo.turris.cz works OK:

DEBUG:src/lib/download.c:78 (download_check_info):Download succesfull (https://repo.turris.cz/hbs/turris1x/lists/pkglists/net_monitoring.lua)

But at the end:

    line not found
    line not found
    line not found
    WARN:src/lib/interpreter.c:130 (interpreter_error_result):Crash report of stack trace dump failed.
    DIE:src/pkgupdate/main.c:179 (main):
    download: Download of https://repo.turris.cz/hbs/turris1x/packages/core/kmod-lib-lzo_5.10.148-1-42186e8efb70da7a8430b969416854a2_powerpc_8548.ipk failed: Couldn't resolve host 'repo.turris.cz'
    Aborted
1 Like

Hmm. Turris 1.0 seems still fine, but Mox doesn’t. Keeps randomly rebooting (at least it doesn’t need to unplug anymore it seems) and also LUCI is missing top menu in like every second time I try to access it.

LuCI top menu got fixed for me by cleaning cookies. (once, fixed permanently)

1 Like

Ha! and just a after this error, taking the link, and wget it , just work.
So :upside_down_face:

Same problem in my Turris Omnia and I need kmod-ipt-nathelper-rtsp for my TV provider :disappointed_relieved:

The version of turris 6.0 Based on the last OpenWrt 21.02.5 release. Kmod-ipt-nathelper-rtsp package to control for example, direct iptv is based on OpenWrt release: OpenWrt-21.02.0 (your nickname sounds familiar to me wireless security forum) so we have to try to update it for the current version of turris 21.02.5 because for the latest version of openwrt 22.03 I see the patch.
I open a issue here: https://github.com/openwrt/packages/issues/19726