Web Squad ISP

Status
Not open for further replies.
Hi all

Has anyone with a Mikrotik added a 4g failover?

I've been looking through guides and see many rely on disabling the dynamic route and using a watchdog on a static gateway.

Would this work or any other recommendations? TIA

 
I'm an idiot. It says "Webafrica" not Web Squad. My brain only registered the "Web" and filled in the rest through association. :sick:

Apologies for the mix up and thanks, now I can safely unsubscribe/block these.
I have been getting a lot of SMS marketing spam from WebAfrica too... How they got my number who knows... someone must have sold my information...
 
If you can get the USB modem to work (this can be a pain at times), there’s two ways to go about this. 1. The simple way, relies on the router detecting that the gateway is down (which you can do if you’re using pppoe) and 2. Using a netwatch script like the one above (which I don’t recommend) and make sure it monitors connectivity to the internet and not just a local interface.

The most common outage will be last mile or middle mile. This will drop a pppoe connection immediately. So setting your LTE’s gateway as a default route with a higher distance than the Pppoe will allow for this route (LTE) to be active once the ppp drops. On Vuma trenched, you can set a static default route (just copy the existing dynamically assigned path and disable default gateway on the dhcp-client) with a ping check on the gateway for the Vuma path and a longer distance path to your LTE’s gateway. For Vuma trenched, and middle mile issue won’t automatically result in a gateway ping check failure as you may still be able to ping the next hop (depending on where you are), so you may have to manually re-route.

The above takes care of routing. Now you need to take care of NAT (determining which outbound IP your device will use to masquerade traffic). The simplest method here is to create a WAN interface list (Interface > interface list) and add both the internet interface (eth1 or pppoe) as well as the LTE interface to it. Make sure your inbound firewall rules reference this WAN interface list and not a specific interface. Then under firewall > NAT, make sure your src-Nat masquerade rule lists the outbound interface list WAN, and not a specific interface. NAT will check for the active route and automatically masquerade out of that interface. When the fibre comes back up, the default route will change and connections will gradually (fairly quickly) move back to the fibre.
Just sacrifice a chicken that you don't experience flaps. Chickens/flaps? I'll get the door. :ROFL:
 
I'm an idiot. It says "Webafrica" not Web Squad. My brain only registered the "Web" and filled in the rest through association. :sick:

Apologies for the mix up and thanks, now I can safely unsubscribe/block these.

No worries. They’ve been spamming a few of our mailboxes too. So at this point I’m pretty sure they’re just scraping emails. There’s another thread here about them not unsubscribing users either.

I guess they’re feeling a little left out after not being able to find a buyer for their business since they went up for sale 2 years ago.
 

Haven’t tested this yet but looks pretty impressive.
 
Nope, some improvements in advertisements and better route selection out of CPT.

Thanks for this btw - feels a lot better, even with the increased latency to some destinations.

Edit: Case in point, best result I've ever gotten to London:

 
Last edited:
Thanks for this btw - feels a lot better, even with the increased latency to some destinations.

Edit: Case in point, best result I've ever gotten to London:


Loving the performance to London as well

Coreix:
11579917190.png



31173:
11579921677.png



Burstfire:
11579934777.png


Seems latencies are a bit confused depending on host. Can't complain about the speeds though.
 
tracert qosping-aws-eu-west-3.ol.epicgames.com 200 vs 150 normally

1623757255032.png


tracert qosping-aws-me-south-1.ol.epicgames.com 300ping vs normally 150

1623757289062.png
 
tracert qosping-aws-eu-west-3.ol.epicgames.com 200 vs 150 normally

View attachment 1089173


tracert qosping-aws-me-south-1.ol.epicgames.com 300ping vs normally 150

View attachment 1089175
These are both on AWS' side, nothing we can do to improve from our end unfortunately. As you can see, we hand over to AWS's network at hop 4. Looks like they're routing east coast at the moment; usually happens when WACS flaps. They drain WACS and then run it in for a few days before switching it back up. I also believe there's a capacity constraint on their WACS route, but this is unconfirmed.
 
thanks for the above,
Will we be getting a communique regarding the speed changes as i might need to/want to upgrade
on the 1st to the new speed.
 
thanks for the above,
Will we be getting a communique regarding the speed changes as i might need to/want to upgrade
on the 1st to the new speed.

Definitely, new packages taking effect 1 July 2021- remember line speeds are set by FNOs and not the ISPs.. That’s why I hinted to them yesterday. Commercials were sent by Vuma late yesterday. Our commercial team need to review and put together the prices and will send out official comms later during next week.
Our commercial team are finalising the new prices and packages, and communications will be sent out as soon as they are done. Seeing as there is a mix of product changes as well as price changes, it takes a few days to complete this process.

Octotel Upgrades have already been processed as well.
 
Speaking of Octotel: There's been a subtle speed bump for all clients on 15 and 30 Mbps packages. You will find that you've got a little more horsepower for free as of today. 15 Mbps > 35 Mbps, and 30 Mbps > 50 Mbps. Sorry, nothing on the larger packages.
 
Our commercial team are finalising the new prices and packages, and communications will be sent out as soon as they are done. Seeing as there is a mix of product changes as well as price changes, it takes a few days to complete this process.

Octotel Upgrades have already been processed as well.

Would you be able to post said updates here as well when it happens. I'm a very lazy email checker.
 
Please use the billing limits, management tools and notifications on cloud providers. Learnt this the hard way on Azure; we set up an expressroute lab for a project and left that running for two weeks while we tested it. We blew past the “free tier” within 2 days and didn’t check the billing (rookie error I admit). Let’s just say accounts wasn’t that impressed.
And then people ask why I'm NOT using AWS/GCP/Azure but OVH?

it's easily a 3rd the price (normal month to month) using OVH services than AWS/etc. not as "slcik" but then, without a huge VC funding, you make use of things that works... basically I'm running a micro PublicCloud as there is a need and I help my clients save money :)
 
Hi, any known issues in Sandton area (Gallo Manor exchange)? I have a flashing blue CPE light. Started about an hour ago. Logged a ticket already.
 
Hi, any known issues in Sandton area (Gallo Manor exchange)? I have a flashing blue CPE light. Started about an hour ago. Logged a ticket already.
Yep everyone in Gallo is down right now.
 
Status
Not open for further replies.
Top
Sign up to the MyBroadband newsletter
X