Been having VDSL issues since Sunday when we lost power… My modem would get an acceptable (relatively speaking) download speed, but the upload was under 1 mbit (usually get 8 mbit upload, but paying for f–king 20!)…
ISP sent out an NBN tech (NBN is the name of the shonky company Murdoch’s cronies in the LNP setup to foist a “solution” of bandaids for our 19th century telecommunications infrastructure - the shambles it is today is 100% due to ideology and nothing to do with technology or $$$) today and he determined that my modem was at fault - as he was able to get 9 mbit upload. Sure enough - I hooked up my old modem, and was able to get 8-9 mbit upload speed.
So - it’s on a different TCP/IP network… Whole bunch of stuff to reconfigure…
My TrueNAS system repeatedly gave back “protocol” errors when trying to mount NFS on the new IP network (from my Pop!_OS desktop - using the CLI)… Took me a couple of hours - but realised (after logging into the ZSH shell in FreeBSD) that /etc/exports had my old VLAN for the shares (had to dig into the TrueNAS web UI to find that section, it’s not under shares, or NFS service!).
That sorted… then my Pi3 running TVHeadend (with a “free to air” TV card for Pi) refuses to let me use MPV to stream ANYTHING! I can’t find any config file anywhere that might have access lists… I think it might be due for a rebuild - maybe with something later than Raspbian Stretch.
My IP cameras? Forget about it… Too hard basket. I reckon I’ll be up for a f–king factory reset!
Any my daughters recently bought a cat toilet with IoT - that will definitely shit (sic) itself… IoT is a massive piece of crap when it goes wrong… hairpulling stuff!
Also - was unable to connect my Android, or my Pop!_OS laptop to the old 5 Ghz WiFi access point - but same password for the 2.4 Ghz was working. Had some non alpha chars in the password - so changed it to all alpha and both Android and Pop!_Os could connect. WTF?
If the old router wasn’t such a shonky toy (Netcomm NF18V) - it might have the option to set an IP address on one of it’s gigabit LAN ports - then I could add a static route (i.e. to from each network) on each device - but no! It’s a TOY!
I have a pretty fair understanding of TCP/IP - and could once quote the OSI reference model, and where every layer “lay”… I’ve been doing subnet masks and CIDR for 25+ years… splitting shonky little Class C subnets into smaller ones longer than Millennials might care to remember…
I know enough that adding another device that has its own DHCP scope will play havoc with everything else…
I have a few rules - NAT mostly - that allow me ingress to my home LAN, and some dedicated lease…
Proof the old router is a “toy” - the dedicated leases are from my old LAN network - this thing is so stupid it doesn’t know when I change my LAN… piece of crap (to paraphrase one of my favourite songwriters : Neil Young)…
That may be what is bugging me.
To my way of thinking, the router should be just another member of the local LAN… but
my Telstra router wont do it… I cant give it an IP address, it wants to give out addresses .
It wants to control everything like a star network.
I can live with it.
OK - I’ve plonked my Gigabyte Brix with a 2nd NIC (USB 3 Gigabit) and will setup OpnSense (BSD based) firewall between to two routers - and - hopefully still use the TP-Link
Public Cloud ↔ Netcomm ↔ OpnSense ↔ TP-Link
Probably just run OpnSense off a 16 GB USB 3 stick… I don’t think it’ll need more space than that…
It was formerly mostly unused running RHEL 8 for doing “stuff”… But if I need RHEL again - I can always spin up a VM…
Update - well that was no good… OpnSense… just kept boot looping once I installed it from USB onto another USB - so gave up.
Trying OpenWRT now… But will probably give up…
I might head into JB-HiFi (electronics store chain downunder) and get my router replaced under warranty - if it can’t go over 1 mbit upload, but the 6 year old device can - it must be faulty…
I’ve just backed up my config - which I can then restore onto the replaced item…
I like Dan’s idea of replacing the newer device because it sure should perform better than the older device.
As far as why use your own router, that’s always my plan too. The WiFi coverage is going to be better than whatever cheap device they provide from the ISP as a default.
The Telco’s router is a “SOHO” toy - very basic and limited.
It’s also 6 years old.
I got fed up of its limitations last year, and went and bought a much later with more features TP-Link branded router. In 2020 when I first started WFH (working from home) during Covid - I discovered a fairly severe limitation - it would only support 16 devices on WiFi - and - somehow we managed to hit that - i.e. Laptops, Macs, iPads and smartphones! I managed to figure out how to boost that to 32 - but - still a PITA.
The WiFi on my “third party” router, from TP-Link is an order of magnitude faster and more resilient than on the ISP supplied device (ISP is #MATE#LETSBEMATES) Netcomm NF18ACV. Everything is more sophisticated.
Had issues on and off with my VDSL since January. The NBN tech pointed the finger at the “new” TP-Link device - i.e. he was able to get 9 mbit upload speed. When I plugged the old “toy” device back in - yeah - I was able to get 8 mbit upload (instead of the sub 1 mbit upload from the TP-Link).
So - if I get it replaced under warranty - and the issue still happens - then - I know it’s an issue with either
my copper
NBN
my ISP
I kinda liked the idea of having a Linux (or BSD) router - but - it’s a lot of work - and - what happens if I’m not here? Who’s going to fix that?
I’m hoping when I get the replacement router back home again - and restore my backup from the “faulty” one onto it - my IP cameras will start working again…
Also - the TP-Link device supports UPNP on it’s NAT rules - i.e. dynamic rules that clients trigger - e.g. my self-hosted cloud sync, ResilioSync, so that it can work from “outside” from the public cloud!
Since we went to fixed wireless have not had one outage. Before that with ADSL on copper and 12k from the exchange we had regular outages with flooding and people digging up cables. Avoid copper if you can.
We are with Telstra. Their fault reporting has improved dramatically compared with what it used to be like to ring up and complain. The free Telstra modem has one good feature… and inbuilt mobile phone backup… if the fixed wireless fails it automatically switches to the mobile phone system and no extra charges!!
Is that with an antenna pointing at a wireless thingie - or - 5G?
I checked one of the ISP’s in Perth to see if I was elligible for Fixed Wireless (i.e. with an antenna on the roof) and they said no…
I think Musk’s “Starlink” is really only suitable for rural areas too… and astronomers HATE those starlink things…
Telstra 5G is only $10 a month more than I’m paying for shonky VDSL - but that 1TB monthly quota looks a bit “limiting” - I probably wouldn’t go over it - but - you never know! I remember running into all sorts of issues when I used to have a monthly quota on my ADSL (my daughter would often cause us to go over - so we’d be shaped down to dial up speed - which was especially painful when I was trying to WFH). But - that Telstra 5G is ~500 mbps download and ~50 mbps upload i.e. 548/52 (vs me paying $75 for 50/20 VDSL, and getting 30/8).
I might head into a Telstra shop today or tomorrow and see what sort of deal I can get for :
5G internet, plus two SIM only plans (my handset and the missus’)
No not 5G… there is an antenna on the roof pointing to a tower on a local hill about 1km away.
You have to have line of sight to the tower.
It is the only option here… no cable, not even fttn.
Needed to fire up my RHEL Brix - so I could verify something - does “dnf update --assumeno” work exactly the same was it does with yum command (for work - doing some doco).*
Plugged it in - powered it on - couldn’t get to it!
Ended up hooking up a monitor and keyboard (NextDock is VERY handy for this) and configured /etc/sysconfig/network-scripts/ifcfg-XXX to change it from DHCP to static assigned…
Rebooted stil not online…
Tried another cable! IT WORKS!
Wasted an hour 'cause of a dodgy UTP cable!
* assumeno is very useful - you can save a transaction log and replay it later - this is very handy if theres a delay between patching two systems, e.g. week 1 you patch test or dev servers, week 2 or 3 you patch prod - this ensures the same package “delta” is installed everywhere… If DNF does save a transaction log, it doesn’t tell you the name or where it might be - so I’m guessing “no” - assume no just says “no” but doesn’t save a transaction log - it’s a dummy run… and yum on RHEL8 is just an alias or a link to dnf…
Here’s how you use that feature on RHEL 8 and later :
You do the update.
Then get the yum history and save as a json file - then copy that json file to the other system you want to get the same patchset.
