This WiFi issue has been there since the very beginning… Seem like no one cares… 
We do care. But instability? What does it means? Are you sure that you don’t have it set on channel with interference? Does all your devices have problem or just some of them? But in reality what can we do. We had some known issues that are now fixed. We do care, it’s ours wifi router at home too (although I for example have no problems at all). But in reality what we can do is just to pray that problem is in opensource part of wifi. We have no chance of getting schematics for those cards or source codes for their firmware. So result is that we know that some of our users have problems with wifi instability. We are trying to found out where problem could be. We are glad that some of our users are also trying (they might have more luck as they are experiencing those problems). But we can’t promise that from some specific release all problems will be fixed, we just don’t know that. With every big Turris update we are looking to upstream for fixes for our wifi stack and applying them, hoping that they are fixing something we missed. One day we or upstream will locate a problem but until then you should know that we care and saying that we don’t makes us angry at you.
3 Likes
“Instability” means that a basic Turris device, out of the box and with no modifications, when automatically updated, stops working reliably. This has happened to me twice.
Immediate forward functionality is the first thing you should test for, and in fact is the cheapest thing to test. You need to discuss this with your QA department.
I’m sure you care but I guarantee you don’t care as much as I did when I was in a different country and my expensive Turris router auto-updated and stopped allowing me to access my home network until I went home and fixed it. I assure you you could not possibly be as angry with me as I was with you on that day.
1 Like
Are we talking about Turris 1.x? That was very unfortunate that it happened. :’(
I think that saying “nobody cares” about the WiFi topic is truly quite too harsh. I also don’t doubt that you guys spend time on determining where the issue comes from.
Well, what does instability imply: Let’s consider my very basic setup that utilizes the default WiFi adapters:
QCA9880 (radio 0):
Mode: AC, channel: 36, width: 20MHz, modus: master, enc: WPA2 PSK (TKIP, CCMP), power: 20dBm
MAC80211 (radio 1):
Mode: N, channel: 6, width: 40MHz, modus: master, enc: WPA2 PSK (TKIP, CCMP), power: 19dBm
As can be seen, there is nothing special. To this, there are no 5 GHz wifis in my environment, while the channel chosen for the 2.4 GHz is not on the same the others are using. I have to admit, channel 6 is typically overloaded, but luckily somehow in my case not.
The clients that are utilizing the WiFi are currently a Lenovo X240, a Lenovo T470s, a Honor 7X and a Samsung A5.
What happens is that the WiFi connection sporadically seems to not transport packets. Although the connection is still on, neither ping nor nmap nor tracepath nor any other low level networking can at this point transmit anything. Also the IP address of the router can’t be accessed with a web browser anymore. But still, the connection is on. Well, yes one can say: WiFi is still running, but neither sending nor receiving.
Only solution is to reboot the router. At least TIME i.e. waiting is not fixing the issue.
Interestingly enough: the network on the LAN ports is still running fine. Full bandwidth, also the LXC container …
Another one and that’s the reason where I’m close to fed up with: I like that the router gets automatically updated. Hence, I’ve turned on this feature. But it seems that each major new update has a 50/50 chance to kill the WiFi, hence requiring a reboot of the router.
I have exactly the same setup regarding the WiFi except more interference and different clients and have no problems whatsoever. We do care, and we test stuff. But if we can’t trigger it, we can’t debug it and then it is hard to fix it.
Exact same bug here myself, word for word!
This wasn’t an issue before (and I’ve had my device since late August). I’ve only noticed these wifi issues since the 3.8.4(?) release which addressed the KRACK bug.
I did create a topic for this issue here which might be of some help for the likes of @miska?
EDIT:
I’ve also tried alternating radios. So I typically use the Generic MAC80211 802.11bgn radio for my 2.4GHz wifi but this has caused issues. Since yesterday, I disabled that radio and used the Qualcomm Atheros QCA9880 802.11bgnac radio in 2.4GHz mode for wifi instead. Still the same issues.
This leads me to believe it isn’t an issue with individual radios.
From the get go (IGG Supporter), the wifi issue was there since I receive this router. I have 3 routers (Asus RT-AC87U and RT-N16) and turned off wifi on all besides the Omnia and only the 5GHz is on. I tested by turning ON 5Ghz on the 87U and turned OFF 5Ghz on the Omnia. I let it run for a month, no issues and no disconnects. So, the 87U can handle the wifi properly. Why can’t the Omnia do the same???
This is what frustrates me and you can’t blame me on this… I just rebooted with 3.9, lets see how this goes…