New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
Honestly, we've found the network to be more stable without quite a few IXs (most of the missing upstreams), which allows our interconnection to work a lot better. We'll likely bring back a few as we find necessary (most likely London), peering obviously becomes more of a challenge without them but peering also can cause an extreme amount of issues we've seen when we were connecting to other providers.
In regard to the T2s, one of them was for the speed of migration. Telstra currently is quoting a 9-month lead time for new ports. We found a solution that will work temporarily which benefits us and our customers. RETN has been on our network for quite some time so not really sure what else to say on that.
Yes, all that request SLA credit during unplanned maintenance receive it, if you didn't and feel you fell under the terms (or didn't, we've handed out SLA for things out of our terms) feel free to reach out to me here, discord, or a tickets, and I'll get that taken care of for you.
I can't confirm this, I'm curious to see if it can be clarified, all my attempts have actually always been ignored
Love how you're trying to sweet-talk your way out of this. Yet Path.net lost over 50 downstreams since July 2024. Says more than enough.
And no, Rust isn’t your only game, but according to your own website, you were "the current top hosts within Rust, with the most players." That dropped fast...
Oh btw, here’s the full graph. Looks like GSK picked up quite a few customers:
You got 180+ Rust servers to change their IP — that’s quite an accomplishment.
Yea dunno what to tell you as whatever I say to you and whatever you say to me won't change the fact the income definitely hasn't gone down 50 customers.
Who wouldn't advertise that fact? We still host quite a few top Rust servers. I really don't know what you are getting at. Got an updated graph for this month post-migration? Would love to see the difference just for curiosity's sake.
I have no idea about you personally, to be honest. We handed out a fairly generous SLA for the migrations, in my opinion.
Hahahah the only thing I've seen so far is a block on discord after I mentioned that royalehostings (actual planned) datacenter migration went a lot smoother.
An SLA credit while for half of that SLA period the servers were offline, unusable, or didn't meet the advertised specifications.
Oh right, routing locations like Singapore over the USA is super useful too, isn't it Kyan?
You specifically were granted the terms of service, not everyone. I don't know why we would go above and beyond for someone who personally messages every person in our discord. Who knows, though, just me.
Tempest and Path have both also pulled off perfectly fine migrations previously, I'm not saying it was a perfect migration this time.
Scroll up a few pages perhaps. Might answer your question I guess? Just a random statement
Hey, former Path.net customer here.
We were with Path.net for about 2-3 years, but eventually we had no real choice other than to move fully to another provider. We actually started that process back in March 2024, and I’ll explain why.
What Kyan is saying doesn’t line up at all with what we experienced. Starting in early 2024, Path.net dropped several major transit providers and exchanges like Telstra, NTT, DE-CIX, and AMS-IX. Instead, they single-homed everything to GTT in Europe. That caused all kinds of packet loss and latency problems. Since it was single-homed, there was no way to reroute traffic. RETN was added later, but honestly it didn’t help much either.
None of these changes helped us. If anything, the issues that were already there before, just kept getting worse. It wasn’t just routing. We saw more problems with filtering, unstable network performance, player disconnects, DNS failures, and outages. It kept happening more and more often. To us it felt like they were cutting costs, not improving the network. That’s what made us decide to start moving away.
Even now, as a former customer, we’re seeing the same thing in the US. Once again they went all-in on GTT, and it already looks like it’s going bad. Their New York location has been down for over five days now, supposedly due to a fiber issue. At least that’s what they’re claiming based on what I see in this thread.
I could go on for a while about everything we dealt with, but I’ll keep it short. Just know that there’s a lot more to it.
We really gave Path.net a fair chance. We had high hopes. But in the end we lost too much business because of the ongoing issues and moving to alternatives was definitely the right call.
What really bothered us was how they treated us during the major outages in January. We were either ignored or lied to. It was by far the most unprofessional behavior I’ve seen from a provider.
I’m pretty sure @dataforest had a similar experience.
Unfortunately most of these issues Jordy listed were issues that only he was reporting for the most part, I don't think we've had any complaints of EU/US ping or DNS issues since the migration (excluding what happened during the EU power outages). We still had Telstra (and will have Telstra once the port is complete). Dropping the other IXs improved the network latency from our testing, maybe you just haven't seen it since.
We are already working on redundant providers for GTT issues like this, but GTT has done us pretty well otherwise; every T1 will have its own issues and to be clear if the second GTT port had been completed (request in for I believe 6-8 weeks now) this wouldn't even be an issue.
I think everyone should make the decision that is best for their own business. Large companies change various things all the time. If you've decided it's best for you, I am glad.
Hard to say how things are now since we moved away from Path entirely after all the downtime and datacenter migrations. But before that, things were just way below what we consider acceptable. The constant issues and major outages in multiple locations were what finally made us pull the plug.
And honestly, I don’t really care if I was the only one reporting ping or DNS issues. That doesn’t change the fact that the problems were there, and they were on your side. I had to deal with the impact and wait for you to fix things every time.
As a former Path customer, I can confirm that we experienced many of the issues that Jordy posted on his discord.
I just didn't bother reporting them all because I figured you would pay attention to a bigger customer and didn't need the same issue reported 10 times... and you were already ignoring or taking weeks to respond to me.
We didn't take part in the DE-CIX shutdown, we were already gone because of regular outages in Frankfurt - all our customers who paid extra for Path.net were already more than pissed off at that point. We passed on SLAs to our customers (we never received them from Path), but Path only made empty promises.
We can confirm the constant issues, there was almost no day without any problems, BGP flaps and the like. Nothing was done about this except empty promises. The only thing that came on time were the bills
I have no idea who you are but as long as support goes through the right lines, it will be responded to. We don't particularly pay attention to anyone more than another when things are done correctly.
If you left before DE-CIX that was entirely different issues than what was described above.
We've also opened communication now since he reached out via DMs.
It went through support.path.net and slack. And while I was waiting for my tickets to get responded to, I noticed your employees talking on frantech discord the entire time.
Mind throwing me a few ticket #s for me to take a look at in DMs?
Let me summarize what @KyanG7 told us so far:
All complaints and criticism re Tempest/Path are merely lies and probably born in a conspiration!
The total 143% truth is that Tempest/Path is a great, super-reliable, and heroic provider!
Any alleged outages are either lies or already are repaired. Our current availability is consistently north of 143%!
It should be added though that Tempest/Path is a very, very modest company, even to the degree that their good work is invisible
Personal remark: unfortunately the credibility of Tempest/Path is pretty much zero. How about actually fixing the cluster-fuck you operate instead of denial, lies, and shadow-fighting, KyanG7?
i've been reading through this and this feels like some high-tier gas lighting and complete denial of data collected and provided. i'm curious to see what the judgements end up being against them from the court filings. this drama is a fun read
No, quite a few complaints are completely valid. The downtime is completely unacceptable, and I am doing what I can to bring it back faster and with redundancy, but when I try to bring that part of it up, it gets slammed down by other completely separate things. I wished to bring Singapore and Tokyo back as soon as we could but as I said above, Telstra has been giving 9 month quotes (feel free to reach out yourself, we've been fighting this fight nearly daily for months) and that is just something we cannot control, there are a few other options in the region but nothing that would have good connectivity in my opinion and if they do, it has the same deployment time issue. I think there has been a lot of misinformation from one side, but also miscommunication from our side.
I don't know where I said our good work was invisible. We defend quite a few "good-cause" companies (although some may argue good-cause is an opinion), I just don't think it's fair to request exact numbers and information that cannot be shared without jeopardizing other important parts. I'd love to share what I can, when I can. I think I've shown that at least with my previous conversations in the LET discord, ArmoredVM, and FranTech that when something CAN be shared, I am fully open to the conversation. If you can find something that is an actual, straight-up lie with evidence, I would love to correct it, so far a few people here have just said no evidence lies and there's simply nothing we can do about it but continue to move on.
Go pay your bills on time and stop lying.
It clear that @KyanG7 is trying to perform some kind of PR safe. He’s bending the conversation to make himself and Path look better. But let’s be honest, he doesn’t care about getting a ticket ID from you. That’s just PR fluff. We’ve had tickets sitting open for weeks or even months, and some were just closed without any actual resolution.
We know exactly what happened. And so far, all we’ve gotten from Kyan are just words. No proof, nothing we can verify. There’s no reason to trust what he’s saying.
To give credit where it’s due, Path’s filtering product is excellent. But the network stability has been terrible. It started going downhill in late 2023. That’s not just my opinion. Other Path customers I’ve talked to have said the same thing, and so have people here like @dataforest and @bobert.
The removal of DE-CIX, AMS-IX, NYIIX, and others was explained as being “better for the network,” but what followed was more instability. It really seems like Path was trying to cut costs. If we hadn’t started looking for alternatives in 2024, we would probably be out of business by now. The weeks/months of outages and instability hit hard in Singapore, Tokyo, Chicago, Miami, New York, San Jose, London, Amsterdam, Frankfurt, Phoenix, Sydney, and more.
And then there’s the Miami outage. From the moment it went down, Kyan was dishonest. The first five days our questions were ignored and the only thing Kyan said was: “This is not Tempest support,” which was incredibly disrespectful. Especially considering we were still running 10 Tempest nodes which we used our Path commitment on and Kyan clearly working for Both companies.
Then on day five, he said the problem was caused by “a haywire machine causing breakers to trip.” That was just a straight-up lie and they were discussing with Equinix (That actaully just cut them off). They already knew the real reason, especially since other locations had already experienced similar issues before Miami went down.
Kyan was asking for a straight-up lie with evidence? Well here you got one of many:
Dropping transit is normal. Dropping IX is normal. pretty sure its common in the industry
but dropping a live pop and blacking out for months is not normal. what happened at the first place for APAC?
telstra doesn't deserve the blame, eventhough im taking your statement with a grain of salt
No, it is normal, just a routine maintenance to ensure stable routing
We migrated Tokyo and Singapore to a new facility which means we need new ports which currently are on a 9 month hold.
Wasn't lying, there was a power issue. I stated above Tempest support needs to go through Tempest tickets, this is the exact reason, they are two separate companies so the engineers of Tempest need to see the Tempest tickets which they can't if its in a Path slack.
Let’s cut the crap, saying Tempest and Path are “separate companies” is just bullshit and you know it. Sure, on paper you’ve made it look that way, but do you seriously think we’re that stupid?
When both my MIA GRE tunnel and the Tempest servers were down at the same time, it was obvious something bigger was going on. Telling me to open a separate ticket with Tempest just to hear “we’re working on it and have no update yet” felt like a slap in the face.
And then there’s the power issue. Just say it, Equinix pulled the plug. You knew that. Pretending it was some rogue machine tripping breakers is just insulting and was a lie.
You straight-up lied and kept us in the dark for days. Then you throw us an SLA like it’s a gift, during which half our servers still didn’t work. That’s not how you treat customers who rely on you.
I expected better from Path. Just be real next time.
Though honestly, bending the truth seems to be a recurring theme at Path.net.
Nah mate, they definitely removed ixps themselfs, de-cix definitely didn't confirm it was because no bills payed, my source at a dc also lied to me about bills not being payed, everyone lies except path.