What's new
Carbonite

South Africa's Top Online Tech Classifieds!
Register a free account today to become a member! (No Under 18's)
Home of C.U.D.

Review Starlink_Part1_draft_final_final2.doc

mikewazar

ENDGAME.CO.ZA
Reseller
VIP Supporter
Rating - 100%
217   0   1
Joined
Nov 26, 2012
Messages
3,992
Reaction score
2,383
Points
10,385
Age
25
Location
Cape Town
I GOT LAZY SO THIS IS PART 1 I DIDNT EVEN PROOF READ PLS IGNORE MY GRAMMAR​


What is it?

Starlink uses many low orbit sattelites to get packets from your router to wherever it needs to go at lower latencies than traditional sattelite based internet because the sattelites are closer to the surface of the planet. The implications of this are that lower latencies make your internet experience feel snappier as the physical time taken for information to travel is greatly reduced, so when you open a website, it will feel more responsive and immediate. Gamers would already be familar with how latencies (pings) can affect your experience but I'd like to explain why this goes further than just your ping in a game.


Starlink has lower latency, and that matters


Latency in itself can be very frustrating to deal with especially in latency sensitive use cases such as games where the difference between 10ms and 100ms can literally be felt and in calls where it seems like you and the person you're talking to keep talking over one another, however latency also plays a role in what we colloqually understand as internet speed. Internet speed is the figure your ISP usually gives you in the form of a download and upload you can expect to achieve on your given internet package. The difference between latency and speed is that latency is the physical time taken for a unit of information to travel (usually measured in milliseconds) while speed is the amount of information that can travel (usually measured in megabits per second) - so while they define seperate things, there is a relationship between them that is important to understand as it will impact your experience no matter what you are doing.



In simple terms, the higher your latency, the worse your connection will feel, for two slightly different reasons:

If you have high latency, you probably have a poor connection which means packetloss. Packetloss in games would probably feel like you're teleporting around the map and your inputs are not registering while in calls it may sound robotic and difficult to hear; this is because services such as games and calls do not resend lost packets because the packets are no longer relevant as the time they were supposed to arrive has passed. These types of services are time sensitive, while services such as downloads or streams are not, however they are still impacted in a different way. Packets in these circumstances do need to be resent because to complete a file download or watch a stream the data has to arrive, and the packets will continue to be resent until they do no matter how many are lost. Due to this, the network gets clogged up resending packets that have been lost and there is less available capacity to send new packets. If you then consider high latency, the time taken to resend each packet will be high compounding the problem. Thus low latency is important for more than you might think, and Starlink has an objective advantage over other satellite connections.




Hardware
  • The Standard kit:
  • Standard is targeted towards the everyday user such as homes
  • It costs $599 and you then own the hardware for life
  • Includes the dish, mounting base, router, as well as the needed long cables (1.8m power-to-router and 15m router-to-dish) giving you a lot of flexibility in mounting the unit
  • Suitable for both fixed installations such as on your roof and portable use thanks to the very nifty automatic stow feature in the Starlink app and detachable cables
  • The dish features automatic aim adjustment so when you turn it on the device uses its motors to orient itself for best signal and continues to adjust its aim for some hours after booting to dial it in as best as it can
  • Even if you have a large open area to get a good view of the sky, things like trees within it's 100 degree field of view can still cause interupptions in signal as the sattelies used by the Starlink network move around, I tried several different ways of getting the best signal including going on my roof where I have a large flat concrete slab to deploy the dish on. The tall palm trees still prevented me from getting 100% signal over the 48hrs I ran the roof mount test for. According to the app, I had 9 minutes of obstruction during this time which is 99.6875% uptime which is impressive and honestly I could not ask for better. I'm still going to persue a better mount using a pole to get above the trees but for now I'm happy.
  • Electronic Phased Array technology lets it change the shape and direction of the beam without much physical movement, apparently this is the first consumer product capable of this
  • P54 rating with self-heating to melt off snow/ice and a very strong build that can withstand hectic wind when mounted properly
  • Operating temperature of -30c to 50c
  • 100 degree Field of View
  • Rated power draw of 50-75Watts which is delivered via a convienent daisychain NOT available on the High Performance model
  • Uses Type F Schuko power connector (@L0phostrix's nightmare) as this model came from Germany
  • Router (Gen2)
    • 3x3 MU-MIMO Dual band WiFi 5 (surprisingly strong throw)
    • No built-in RJ45 port (I have the $25 adaptor giving you a single port, which you can then connect to a switch)
    • Operating temperature of -30c to 50c
    • IP54 rating but not meant for outdoor deployment
  • The High Performance Kit (brief info just for awareness)
  • It's targeted towards more demanding use cases such as businesses or very rough areas such as the East Rand (shudders...)
  • $2500 (over 4x more expensive)
  • Not nessesarily better if you just want better speeds or lower latency
  • The main limiting factors in order of likelyhood is your subscription plan, satellite density in your area and lastly the model of kit you have
  • Probably only worth the extra money if you live in an area with extreme climate or need the flat design to mount on a vehicle or boat
  • It does have a broader field of view (140 degrees versus 100 degrees)
  • Worse if you will need to transport it as it is larger and heavier and overall not designed for portability the way the standard kit is as you need seperate power for each component
  • If you get the high performance model you will need to pay more for your subscription for network priority else you will be limited by your plan and not the very expensive hardware



Subscription



While the hardware is owned for life, you need to have a subscription to Starlink to actually use it. It does not require a contract so you can pause/unpause the subscription as needed, however due to it not being officially launched here you need to jump through some hoops to subscribe and have a plan to get around the two month roaming limit which requires you return the unit to the country the subscription is tied to otherwise the device will cease working. This limit is not currently enforced however just like in Turkey or Ukraine, Starlink has shown no reserve in shutting down the service with no prior warning.


To get the lowest possible monthly cost at the moment requires you to subscribe in a supported country in Africa to use the mobile regional plan with roaming as it allows use anywhere on the continent of subscription and the sattelite density seems to be more than capable of serving everyone on the cheapest tier without any noticable performance issues at the moment. This may change as more users onboard the network and prioritisation actually kicks in.


Plans available to me within the app at time of writing:

  • Mobile (Regional) allows use anywhere on your continent at best effort speeds without limit (this is the plan I am on)
  • Mobile (Global) allows use anywhere with coverage at best effort speeds without limit
  • Mobile (Priority) allows priority use anywhere with coverage charged in blocks starting from 50GB - 5TB after which you revert back to unlimited at best effort speeds
  • Priority allows use in one fixed location at the highest possible speeds and includes a public IP and premiums support charged in blocks from 40GB - 6TB after which you revert to unlimited at standard priority
I chose to not include pricing as it is dependant on the region you choose. Google should have specifics if you'd like to know more.


Testing methodology

  • DNS servers were set to Starlink default unless I needed to use a DNS proxy to access a particular service during testing
  • Testing was all performed over a wired connection unless specified where I used the included router's WiFi
  • Testing was all performed under load with scripts running various bandwidth intensive tasks throughout thus all results reflect a hell scenario unless stopped for the purpose of a test
  • I was connected to a base station in Nigeria because thats the only place in Africa with a base station



Streaming


Apple TV, Netflix, YouTube, Crunchyroll, Shudder, Prime Video, and F1TV all performed without issue at the maximum resolution available running at least 2 concurrent streams for more than 30 minutes without buffering tested at random times throughout the period.



Porn


The Hub, Xvideos, Porn Dude, and Xhamster all performed without issue at the maximum available resolution running each test about 3-5 times a day without buffering. I ran these tests over WiFi on my phone as testing on my designated machine would have been inappropriate, however this did not impact performance due to the strong signal provided by Starlink's stock router.



HTTP(S) Downloads


Several downloads were tested from my seedboxes in Germany, France, Romania, and Finland and performed within the expected range of 30-200Mbps but note when staring at the download rate it did fluctuate and at times stopped and resumed. Overall this metric should be looked at in the context of the overall amount of traffic I managed to push during the period alongside the below graphs showing the throughput over the entire period of testing.


An iOS IPSW file was pulled from Apple's CDN and performed consistently with the above.


A Windows ISO was pulled from Microsoft's CDN and performed consistently with the above.


Using a multi-threaded download tool (Folx) I spammed connections which gave me more consistent speeds near 150Mbps which leads me to believe the fluctuations are a result of packetloss. More details about this below.



Torrents


I pulled some Ubuntu ISO torrents and performance was lower than what I found during HTTP downloads achieving 10-120Mbps, take note I did not manage to open any ports to connect to peers which may have impacted the performance but this reflects reality. The more costly Priority plan would give you a public IP you can then open ports for to connect to other peers, a VPS would also allow you to open ports which I did also test below. Please take note that unlike South African ISPs that do not care about DMCA and copyright complaints, Starlink will forward you complaints they recieve and ask you to stop - I don't know if this would lead to account termination as I don't partake in piracy.





Xbox


I downloaded a game from Game Pass and it was very poor with wild fluctuations between a few Mbps and topping 70-80Mbps for brief periods which I suspect is due to the CDN I was pulling from having extremely high latency as I found routes sometimes took strange paths through Asia (more info on this strange behaviour below). I stopped/restarted/rebooted a few times to try and see if that changed anything but it did not, I suspect this has to do with the routing from the base station in Mozambique as normally I would route via the west to Europe.





Playstation


I downloaded a game and it performed much better than on the Xbox achieving pretty stable speeds around 90-150Mbps.





Whatsapp and Facetime


I tested voice and video calls to people in SA, UK, and USA



SA - Latency was surprisingly poor. It felt as if there was a solid half second of delay but video quality was good​
UK - Latency was much better and indistinguishable from a call made over my fibre and video quality was good​
USA - Again, latency was much better and video quality was good. I believe the SA tests were routing me from SA to Europe and back causing the extra leg of delay​
This behaviour aligned with what I was seeing in other tests, more details below​



Performance testing over Tor

Even after forcing entry and exit nodes (which I do not recommend) Tor performed poorly due to the latency both to clearnet and hidden services. This is partly due to how Tor works, and partly due to the latency introduced by Starlink having to haul up to Nigeria using lasers before even reaching an internet exchange.


l73Ej5x.png



7VRluC6.png


MHpd1Hi.png



7FV7PbG.png


9bAxlWO.png



AdDOJeH.png


hnvC4aq.png




XzhjY9Q.png



p7wOfZ7.png



qBAx1i1.png





So far things look pretty good and I really don't have any complaints about the performance for what it is considering it takes two minutes to setup, can fit in a boot to move around, and is a very viable option for those without fixed line or cellular available. For the average Joe this is super simple and out of box it just works. It is a shame that officially its not available here, but I hope that changes.


One complaint I do have is due to the lack of base stations in the country. Latency would be greatly improved if we didn't have to route through neighboring countries to reach services within South Africa. Counterintuitively resources located overseas perform better than resources located locally. This is no fault of Starlink as I'm sure given the green light we would have them in Gauteng and Western Cape. Viva ANC.

We would actually be a great place for base stations considering the fact that we are the only country in Africa that can provide routes to both the East and West via Western Cape and Kwazlulu Natal... but anyway



Part 2 to follow... including how I found a vulnerability in the network and got contacted by a Starlink employee at like 3am[/SPOILER]
 
yolo im not fucking finishing this


Digging into Space Exploration Technologies Corporation





Their network has almost 250 000 IPv4 IPs assigned to it so you could say its kinda chonky, looking into how this is utilised I found that only ±10 000 IPv4 were really doing anything you could detect externally with half of them being part of the USA segment of the network. Since Starlink uses CGNAT for their customer IPv4 connections (unless you're on the business focused package) it does make sense to see this as all the customer connections would be using the unholy amount of IPv6 Starlink has which is great to see. I did notice something a bit strange however.



The v6 portion of the network (yes I know geolocation is basically meaningless but read) is basically USA (to serve the North Americas), Chile (to serve the South Americas), UK (for basically the entire region of that hemisphere), and Nigeria (to serve Africa and the islands). I found this strange because the v4 network is quite different in that it includes more countries and it actually matters because when doing some testing, over v4 I am served through Mozambique but over v6 I am served through Nigeria. I'm not sure why this is, but it made research and testing more complicated because I cannot control whether I would connect over v4 or v6 and it changed the results of the test - this was quite frustrating and not consistent enough to reproduce results in a meaningful manner as one attempt would give me a good time and the next I would get a polar opposite result.



Not to be deterred I decided to try using tunnels to force traffic to go through specific protocols and endpoints to try and standardise as much as possible. It actually worked pretty well which enabled some pretty neat trickery I'll expand on further down.





The IPs allocated to my router

Network: SpaceX Starlink AS14593
IPv4: 102.215.58.62 (from 102.215.58.0/24)
IPv6: 2c0f:2a80:1c0:b300:35a2:8b19:15fb:944a (from 2c0f:2a80::/38)
DNS: 138.199.63.129 [ECS]
DNS: 2a02:6ea0:1a03::1 [ECS]


Some interested numbers for the nerds using bgp.tools


IPv4 End To End: 190.4ms
IPv4 TCP Stack: 184.8ms [+/- 23.3ms]
IPv4 TCP MSS: 1448b
IPv6 End To End: 198.4ms
IPv6 TCP Stack: 186.1ms [+/- 8.2ms]
IPv6 TCP MSS: 1428b


The public facing network is pretty much just customer CGNAT and a load of SNMP appliances monitoring everything. Not much else but I did see some unpatched devices:

SMBv3 Remote Code Execution - 10
CVE-2022-32548 - 7
Heartbleed - 2
CVE-2021-31206 - 2
FREAK - 1



As usual, lots of Mikrotiks, Cisco stuff, some Sonics, and everything else looks private. I did find a bunch of customer devices that were running their dishes with their own hardware using Starlink's bypass mode (more on this later) such as this one: http://149.19.173.165:1024/webpages/login.html

Feel free to play around I guess. Moving on.


I hate being bipolar, it's awesome


A challenge then. I decided to fuck around a bit to see if there was anything cool I could do or if there was any room for improvement on my poverty spec hardware and SASSA grant subscription. I began by trying to figure out what Starlink uses for actual connectivity because while they do build and launch their own stuff for the network that runs in space, to actually reach the internet Starlink relies on the same carriers and companies your cellular or fibre ISP uses. Now while Starlink is testing satellite-to-satellite networking, which has the potential to reduce the reliance on these common carriers, right now you actually only connect to a satellite to get your packets from your router to a base station from which the packets use the same boring undersea fibre cables that seem to break every 2-3 years. A quick throwback to my threads from when I was 21:





This might be important because this means that Starlink is NOT safe from these breaks impacting their network despite the belief some schizos on Twitter I fight with daily (@mikewazar on X.com) seem to have. If a cable break happens again, the network is just as fucked as anyone else until satellite-to-satellite communication is officially launched. Keep this in mind when people try shilling you this magic stardust internet solution.


Looking at the range I was assigned I found that over v4 I would either get carried by Liquid or WIOCC which are pretty commonly used by popular ISPs such as Afrihost and Vodacom amoung others. A traceroute to a server in Portugal took me via Nigeria on WIOCC while a trace to Germany took me via Nigeria on Liquid. Looking a bit deeper at why I was taken via Nigeria I noticed that the reverse DNS on my v4 IP actually said Lagos, Nigeria even though the geolocation was Maputu, Mozambique.



Code:
customer.lgosnga1.pop.starlinkisp.net


Over v6 this was also consistent as I was always routed through Nigeria on Liquid as far as I could tell which makes me wonder if Starlink is using Nigeria as their African point of convergence because South Africa has been so hostile to them. Looking at Starlink's PeeringDB info basically confirmed this as there was only one location in Africa that Starlink was using for upstream connectivity.


Nigeria seems pretty well connected so I guess this was a no brainer. The country is supported by:


2Africa
Africa Coast to Europe (ACE)
Equiano
Glo-1
MainOne
Nigeria Cameroon Submarine Cable System (NCSCS)
SAT-3/WASC
West Africa Cable System (WACS)



So I stand corrected when I thought I was being routed through Mozambique. I'm still not sure why the geolocation says Maputu but anyway, Nigeria currently has two ground stations as of writing this post with one of them being convienently identified by this Twitter account: https://x.com/Megaconstellati/status/1658364726982615041?s=20


The other is in a location called Lekki which is in the south-east region of Nigeria, close to the coast where all the cables land according to the illustration above. This is a bit worrying as having the entire continent connect via two ground stations in Nigeria which are pretty close to one another is a consolidation risk to the network. I hope this is improved quickly. A cool resource I found with info on all the ground stations can be found here: Starlink Ground Station: Overview And Locations

Traces from large carriers to my Starlink gateway (Reverse paths)

ZA to ZA
Code:
traceroute to 2c0f:2a80:19e:2d00::1 (2c0f:2a80:19e:2d00::1), 30 hops max, 80 byte packets
 1  gateway (2c0f:e9a0:1:1::1)  1.129 ms  1.139 ms
 2  * *
 3  be2389.ccr22.lon01.atlas.cogentco.com (2001:550:0:1000::9a36:50c9)  185.566 ms *
 4  * *
 5  * *
 6  * *
 7  * *
 8  be3268.rcr71.lis01.atlas.cogentco.com (2001:550:0:1000::9a36:392a)  223.037 ms  233.673 ms
 9  * *
10  * *
11  * *
12  2a01:410:c400:6::3 (2a01:410:c400:6::3)  381.460 ms  384.348 ms
13  2a01:410:c300:10:8000::22 (2a01:410:c300:10:8000::22)  380.159 ms  382.665 ms
14  host.starlinkisp.net (2620:134:b0ff::15d)  393.312 ms  383.182 ms
15  host.starlinkisp.net (2620:134:b0fe:249::23)  403.247 ms  403.163 ms


DE to ZA
Code:
traceroute to 2c0f:2a80:19e:2d00::1 (2c0f:2a80:19e:2d00::1), 30 hops max, 80 byte packets
 1  gateway (2001:978:1:320::1)  0.951 ms  1.016 ms
 2  te0-0-0-0.agr12.muc03.atlas.cogentco.com (2001:550:0:1000::9a19:7e5)  0.972 ms te0-0-0-0.agr11.muc03.atlas.cogentco.com (2001:550:0:1000::9a19:7e1)  1.067 ms
 3  be3244.ccr22.muc03.atlas.cogentco.com (2001:550:0:1000::8275:320d)  1.499 ms be3245.ccr22.muc03.atlas.cogentco.com (2001:550:0:1000::8275:3215)  1.595 ms
 4  * *
 5  * *
 6  * *
 7  * *
 8  be3268.rcr71.lis01.atlas.cogentco.com (2001:550:0:1000::9a36:392a)  45.419 ms  45.424 ms
 9  * *
10  2a01:410:c400:a:41:84:151:63 (2a01:410:c400:a:41:84:151:63)  214.998 ms *
11  * *
12  2a01:410:c400:6::3 (2a01:410:c400:6::3)  215.727 ms  215.740 ms
13  2a01:410:c300:10:8000::22 (2a01:410:c300:10:8000::22)  214.782 ms  214.699 ms
14  host.starlinkisp.net (2620:134:b0ff::15d)  214.738 ms  214.819 ms
15  host.starlinkisp.net (2620:134:b0fe:249::23)  214.676 ms  214.672 ms


NY, USA to ZA
Code:
traceroute to 2c0f:2a80:19e:2d00::1 (2c0f:2a80:19e:2d00::1), 30 hops max, 80 byte packets
 1  gateway (2001:550:1:316::1)  0.812 ms  0.798 ms
 2  * *
 3  * *
 4  * *
 5  * *
10  2a01:410:c400:6::3 (2a01:410:c400:6::3)  264.614 ms  264.630 ms
11  2a01:410:c300:10:8000::22 (2a01:410:c300:10:8000::22)  263.757 ms  263.729 ms
12  host.starlinkisp.net (2620:134:b0ff::15d)  263.549 ms  263.559 ms
13  host.starlinkisp.net (2620:134:b0fe:249::23)  263.702 ms  263.569 ms

SG to ZA
Code:
traceroute to 2c0f:2a80:19e:2d00::1 (2c0f:2a80:19e:2d00::1), 30 hops max, 80 byte packets
 1  gateway (2001:550:1:32a::1)  1.147 ms  1.186 ms
 2  te0-0-0-19.nr01.b019922-0.sin01.atlas.cogentco.com (2001:550:0:1000::9a18:4b81)  1.470 ms  1.577 ms
 3  te0-0-0-11.agr02.sin01.atlas.cogentco.com (2001:550:0:1000::9a18:3d49)  1.671 ms te0-0-0-23.agr02.sin01.atlas.cogentco.com (2001:550:0:1000::9a18:3d4d)  1.492 ms
 8  be3268.rcr71.lis01.atlas.cogentco.com (2001:550:0:1000::9a36:392a)  166.656 ms  170.520 ms
 9  2001:978:2:32::5c:2 (2001:978:2:32::5c:2)  291.330 ms *
10  2a01:410:c400:a:41:84:151:63 (2a01:410:c400:a:41:84:151:63)  341.749 ms  346.560 ms
11  * 2a01:410:c400:5::2 (2a01:410:c400:5::2)  341.516 ms
12  2a01:410:c400:6::3 (2a01:410:c400:6::3)  342.381 ms  347.287 ms
13  2a01:410:c300:10:8000::22 (2a01:410:c300:10:8000::22)  351.705 ms  351.705 ms
14  host.starlinkisp.net (2620:134:b0ff::15d)  337.766 ms *
15  host.starlinkisp.net (2620:134:b0fe:249::23)  337.684 ms  337.646 ms





If you would like to test more:

Target IPv4: Due to CGNAT being used this isn't reliable
Target IPv6: 2c0f:2a80:19e:2d00::1

 
Last edited:

Privacy


While a wireless connection is pretty neat it does come with some implications that you should be aware of. Firstly, your radio footprint makes you stick out like a sore thumb to anyone with a spectrum analyser, meaning your physical location can be tracked pretty easily and to quite a shocking degree of precision if you're in an area with few other users. Keep this in mind because regulatory bodies really don't like it when you're using illegal equipment and spectrum. You can also be tracked directly by Starlink as their satellites can pinpoint your dish's location using the Phased Array Antenna to a precision that might make you think twice about using it. Unfortunately this is the same method used to orient the dish correctly so its a double-edged sword.


Besides physical tracking, concerns have been raised around the protection of your data in transit to and from the satellite itself, with some worrying about the risk of spoofing your dish to intercept communications. While this is mostly a non-issue thanks to TLS, protocols that do not have this luxury need to be considered inline with your particular threat model. Starlink claims all communication to and from your dish is encrypted but very little information about this is available.
 
What are the repercussions if ICSA catch you using this? Not talking personal use, in a commercial space.
 
What are the repercussions if ICSA catch you using this? Not talking personal use, in a commercial space.

They will put you in a cage.

NOTE TO READERS. THIS ENTIRE THREAD WAS A HALLUCINATION I DO NOT OWN A STARLINK DEVICE I WAS ON A LOT OF DRUGS WHEN I WROTE THIS THIS WAS ALL MADE UP DO NOT TRY AND PUT ME IN A CAGE ITS NOT REAL I HAVE BROKEN NO LAWS BECAUSE THE DRUGS WERE PRESCRIPTION
 
They will put you in a cage.

NOTE TO READERS. THIS ENTIRE THREAD WAS A HALLUCINATION I DO NOT OWN A STARLINK DEVICE I WAS ON A LOT OF DRUGS WHEN I WROTE THIS THIS WAS ALL MADE UP DO NOT TRY AND PUT ME IN A CAGE ITS NOT REAL I HAVE BROKEN NO LAWS BECAUSE THE DRUGS WERE PRESCRIPTION

@SpaceWalker I think you going to need a reseller sub there bra if you want to deal on Carbonite!
 
Porn


The Hub, Xvideos, Porn Dude, and Xhamster all performed without issue at the maximum available resolution running each test about 3-5 times a day without buffering. I ran these tests over WiFi on my phone as testing on my designated machine would have been inappropriate, however this did not impact performance due to the strong signal provided by Starlink's stock router.

Dang, 3-5 times a day across 4 different porn sites? I truly respect the dedication to the testing methodology here 😂

Very cool read though, thanks for sharing!
 
This was great reading and very informative, thanks a ton @mikewazar

Any update on this feverdream of yours?

Someone I know is looking to service their people (probably about 20-30 under one roof), which means the Standard/residential unit won't fly due to the restrictions placed on amount of devices that may connect to the SL router.

Does anyone have any experience with the High performance unit?
Do you also have to buy a Wall mount extra?
 
Someone I know is looking to service their people (probably about 20-30 under one roof), which means the Standard/residential unit won't fly due to the restrictions placed on amount of devices that may connect to the SL router.
Just bypass it and connect a beefy SME grade router and it will be fine.

Does anyone have any experience with the High performance unit?
Do you also have to buy a Wall mount extra?
Yes, no different as we are limited by the lack of ground stations.
You can get creative mounting it, consult an expert.
 
Thanks a lot for the quick feedback.

One more thing, how often do you have service interruptions on average - and how long have they been?
Looking to get a rough idea of uptime.
 
Thanks a lot for the quick feedback.

One more thing, how often do you have service interruptions on average - and how long have they been?
Looking to get a rough idea of uptime.

I included my anecdotal experience with graphing, logging, etc above so you can see the raw data in a very easy to interpret manner. Check the original posts.

I didn't ever notice an interruption, and obviously the better your install the better your experience will be. We have enough sats above us that you should not ever have an outage unless you're in a specific region: South Africa - SKA Observatory
 

Users who are viewing this thread

Back
Top Bottom