one month passed, and I see lots of work on version 7 that feels regrettably more as a not reliable RC than something to work with. Since the openWRT is almost outdated, should we discuss to move quickly to 8?
Seeing forward if an discussion arises, if not thanks to developers anyway.
Vienna
Except for Transmission, if you only used Reforis to configure Omnia, there should be no problems with the update. Most of the other problems arise from the fact that users have something more configured.
Yes I only used Reforis - Luci only for DynDNS and Adblocking.
Still having regularly crashes - WiFi, ReForis, LuCi. And is it really reliable that only updates work, not a clean system? Tried to install medkit three timesā¦
But if everyone shares your point, there must be an issue with my omnia.
I had my problems as well with a Turris MOX which fails every main or major update. Going via clean install solves part of the problems, but others need to get solved in a 7.0.x version (Opkg in stable 7.0.0 uses hbl instead of hbs).
But just as @peci1 wrote there are multiple possible configurations that will just thwart the update or get thwarted by the update. The main developer of AsusWrt once told me (in 2017 or 2018 if I remember correctly), the TOS project is doomed because such updates cannot work. Well - I still own my first TO and bought 3 others + 2 MOX in the mean time, because the updates work that well. OpenWrt updates only seldomely worked for me and never smoothly. Turris team needs to get as much as possible as fast as possible upstreamed so they have more time for supporting the non-OpenWrt partā¦
Yet a 8.0 is right now not possible - OpenWrt 23.03 is based on nftables and as Turris team (wisely - this would have multiplied the problems observed by users) decided to separate version updates from functional updates, we need to get 7.1 with nftables (aka fw4) first, before they can update to TOS 8.0/OpenWrt 23.03.
THX, but this Turris Omnia was bought 2016 and has not been in use for more than 3 years. It has only uptime for less than three years, normally I would not expect this.
root@L*:~# mmc extcsd read /dev/mmcblk1 | grep LIFE
-ash: mmc: not found
root@L*:~# mmc extcsd read /dev/mmcblk1 | grep EOL
-ash: mmc: not found
@ssdnvv You might failed to read that I tried this three timesā¦ I made the update with a minimum config to avoid this.
I also had my problems to activate a new wifi card, which should be supported since 6.0.4 simply becouse a package was missing. [OpenWrt Wiki] package: kmod-mt76-sdio - did the trick for me.
But isnĀ“t that an argument for faster forwarding in order to keep more up to date?
And reForis does not show actual NTP, which is nothing big itself, but keeps me reminding to check everything. So I spent some time yesterday to get a glimpse of my errors - while browsing around I had two more events that made a restart necessary and I could not find out a lot, since once happend while instrumenting LuCi the other while using reforis. Provided the diag files, which took always quite a bit - nearly half an hour - to get.
Anyway I thought I was not the first or only one - TOS 7 release date and cruel future - and there should be some discussion within a project how to proceed. For upcoming devices like Enterprise version, which would be a new highlight for me, should it be a strong basis, rather a revolution than an evolution? Not mine to decide, but I am curious to find out.
At the moment spent some time investigating my own device and finding a solution within the project, regretably I am running low right now and I have a device that stops workingā¦
I think one first needs to install the necessary packages, as they seem not to be part of the default setā¦ opkg update ; opkg install mmc-utils
should do the trickā¦
mmc extcsd read /dev/mmcblk0 | grep LIFE
eMMC Life Time Estimation A [EXT_CSD_DEVICE_LIFE_TIME_EST_TYP_A]: 0x01
eMMC Life Time Estimation B [EXT_CSD_DEVICE_LIFE_TIME_EST_TYP_B]: 0x01
mmc extcsd read /dev/mmcblk0 | grep EOL
eMMC Pre EOL information [EXT_CSD_PRE_EOL_INFO]: 0x01
ok, might be nice if we could pin this info somewhere. Thanks!
But @peci1 this can not be an issue, but thanks for the reminder.
version 7 that feels regrettably more as a not reliable RC
Just as a datapoint: I have found v7.0 reliable. No crashes, no special problems. And have updated a couple of times with no problems.
But I am just using it as a secure router, not doing anything special.
Though maybe after I send this it will crash for the 1st time.
if you did a clean install, definitely check if the router can access the packages list and does finish the updates/install any aditional packages. if not, it might be failing the signature verification! my āquick and dirtyā workaround is switching to hbt branchā¦
another reason for unstable operation might be a failing power supply. some wifi cards draw a lot of power, depending on the mode. if the psu is weak, it might give in when thereĀ“s a lot of load, not causing a restart, but errors.
There might be something that I have to wait for in Turris 8 - 23.05. There is a 32-bit RISC MCU - Filogic 630 - build in my wifi, that will be supported there.
@n8mahr Thx, but I eliminated the update issue through medkit install of 7.0.1. PSU and Powerconsumption is no issue.
special thanks to Turris team @TomasZak - I donĀ“t know what they had to fix in my omnia, but set up DynDNS, DAWN, WiFi, AdBlock with brand new 7.0.1 - and it seems to work! Might be different uboot?
My thanks also go to @kai and all here that tried to help me out!