Hetzner has a reputation for locking accounts for "identity verification" (Google "hetzner kyc" or "hetzner identity verification"). Might be worthwhile to go through a reseller just to avoid downtime like that.
Thanks for mentioning Entrywan, they look great from what I can tell on their site. Have you used their services? If so, I'm curious about your experiences with them.
At work, we upgrade when there's a compelling reason (structured logging in 1.21 came close; we haven't finished migrating from our old logger yet). Before deploying that version, we typically bump versions in all our Go repos and run the tests in those repos. Having the same version across apps gives some people on our team a bit of comfort (with Go, it's not as important as some other runtimes).
Came here to say the exact same thing. chi has been reallly great for a couple projects--in fact, it's easy to forget it's even here. Moving them into the stdlib means they'll always be maintained and ensure that approach is used in many Go programs.
One technique is to never upgrade clusters. Instead, create a new cluster, apply manifests, then point your DNS or load balancers to the new one.
That technique won't work with every kind of architecture, but it works with those that are designed with the "immutable infrastructure" approach in mind.
There's a good comment in this thread about not having your essential services like vault inside of kubernetes.
Any state that a container uses, such as databases or static assets, should be mapped to something outside k8s, no? I thought container orchestration was only for app later
> Native Americans routinely burned the landscape—to foster the growth of useful plants, to clear space for farming, and to improve the conditions for hunting.
> ...
> In addition to maintaining parklike conditions, these managed blazes prevented fuel from building up, and so staved off larger, potentially unmanageable conflagrations.
Much of the world still operates like this. Check out the Chiang Mai, Thailand burning season[1].
The US and Canada are some of the only countries where wood is the primary building material. In the rest of the world, stone is used which doesn't catch fire so easily. That might help explain the fear of fires in the US.
The government of Russia, home to the world's largest forest reserves, is trying to subsidize wood-frame buildings. Most modern buildings are made from stone, though, because of fire regulations[1].
The author describes the alerting and notification workflow really well.
We use IRC at work for this purpose. Prometheus alerts flow into channels that anyone is free to join or leave depending on what they're working on at the given moment.
Same here. ZNC on Debian (apt install znc) on an Entrywan instance.
The freenode debacle impacted a handful of my favorite channels. #tcl is noticably quieter and a couple others are still trying to decide whether they're an official channel or not (this affects #channel vs ##channel naming). I'd bet that the number of active users has dropped but I don't have any emperical evidence.
If anyone is considering ZNC, here are a couple suggestions to change to the default config:
1. AllowWeb = false (only enable the web interface if you need it)
2. For each channel, add Buffer = 10000 so that you can have history for more than a day or two (the default is quite small)
Less than half the activity before/after according to this non-scientific analysis (my client wasn't always connected and it looks like the logfiles contain topic/names and other output).
> For example, a bundle may provide 200 minutes of calling during a 7-day period for a discounted price. To avoid losing unused minutes, we learned that as people near the end of their bundle’s time period, some use available minutes to call back unusual numbers in their incoming call-log, which they had not answered.
This probably explains the reason for the accidental social ties.
I lived in a country where mobile phone calls were expensive relative to wages. Much of the time you'd receive a call that only rang once. The caller would hang up after only a single ring. This signalled they wanted to talk to you but didn't have enough "talk time" left. If the caller was a contractor or someone on your payroll, they would almost always employ this tactic to keep their costs down.
When I was a student (around 1998/99) I had to buy a mobile phone, but could not afford a subscription.
Fortunately we only pay for outgoing calls in my country, otherwise it would have been too expensive.
I bought a prepaid sim card, that had an expensive cost per minute (almost $1). I would call my parents landline once, as a signal to call me back.
It's one of the main reasons why SMS was the preferred communication methods between students, because calls were too expensive. This habit continued when we grew up, and now almost nobody calls each other. I guess the phone companies didn't think of that scenario...
I don't get it. If the collect call from A to B is accepted, B pays the cost of the call. In your scenario B instead calls A directly... and B still pays the cost. Where's the savings here?
(And there's a pretty clear downside to this. What if it's an actual emergency from a different phone than the standard. E.g. the only time I can remember answering a collect call, it was my sister calling from a foreign hospital after a serious accident. Trying to call back at a different number wouldn't have been great. I guess you could have a protocol involving multiple collect calls, and hope that there's no record that the first one was rejected.)
I made a collect call (~several states away) home when I was a kid and didn't have a cell phone. I forgot all about that. I feel old now. I remember my dad had to accept and kept it short to reduce costs even though I'd been away for weeks.
Back in the day ~23 years ago. A telco that had just launched started providing voice mail service. Normal outgoing call rates were about .5 usd per minute. The newly launched voice mail service was free. So university students found a new trick to call their friends, flash your friend (single ring signal to say that they don’t pick up their phone), hang up and call again, let it ring till it goes to voicemail, leave a long message, hang up and wait for the reply. Obviously it’s not full duplex and latency was high but what the e heck it’s free. The telco killed it within 2 days… and then we started having fun with all those smsc Center numbers from all over the world…
I forgot about it. Yeah. I had my first cell phone around 2000 and everyone was doing that. In my country outgoing calls were free, charges to the caller were only applied after receiver accepted (answered) the call.
I remember some of my friends being "famous" for doing that all the time for every single call, operating the mobile at like 5$/year. The whole thing had it own jargon.
in some places this works even when the budget literally zero. that is without any money i can call someone, and let it ring until i get the message that i don't have enough money. but the receiver gets the ring and can call back. it makes sense, because it will encourage someone to spend money to make a call.
Since they raised $61 million (source https://techcrunch.com/2024/02/05/cloud-native-container-man...) they probably had way more team than $10 million could support. Since they raised the VC money I guess that downsizing to a team that the business can sustain is not an option.
According to LinkedIn they were 50-200 employees which after excluding all the other costs for running a company, is definitely not enough to cover the fully-burdened payroll of even a 50 person organization that’s probably predominantly Engineering types, that assumption being if most people are US-based (or if they don’t do “location based” pay, and indexed off a more expensive location).
Right. Flux was a handy little tool[1] that sync'd yaml manifests in git repos to live clusters. The concept was fascinating, and the tool was well done--small and efficient. Easy to learn.
In 2019, they announced they'd be "merging" with argocd[2]. It seems the merge never really took place, and after that they deprecated flux and announced flux2[3].
The sudden changes of course were a little confusing and perhaps not too well communicated.
I was hired to support Flux v1 two years after the events you described, in the beginning of 2021 to go on supporting Flux v1 until we could get everyone off the boat.
(I worked at Weaveworks until last month, and I'm still a Flux maintainer! Keep the Flux talk in Present tense please! ;-)
