150 karma · joined November 14, 2016
> After the expansion completes, old blocks remain with their old data-to-parity ratio (e.g. 5-wide RAIDZ2, has 3 data to 2 parity), but distributed among the larger set of disks. New blocks will be written with the new data-to-parity
If you receive 10K USD as income one month, you would pay tax on that and then be able to buy things
If your business received 10K USD it could buy all sorts of things before declaring a profit and paying tax only on the profit.
How is America behind on this? Everything is hosted there and cached there. Is it big business lobbying politicians to prevent them from having to compete? In South Africa we have dozens on ISPs and as an economy we're probably not far off from one US state. The competition has been great for consumers.
On the one hand you have many saying that the aircraft carrier might go the way if the battleship - more and more redundant in a world with landbase anti ship missile systems.
On the other hand there are some saying that ships will envolve to meet the threat.
Pulled the trigger on that three times now.
My first cluster was Raspis and ODroids. The two issues I had - not everything runs on Arm (yet) and memory constraints were tough to work with.
My second cluster - 4 Atomic Pis. The (Intel Atom) quad cores were fine but the 2GB of memory was a real problem when I started playing with more interesting workloads.
The current "production" home cluster is 4 old laptops. They're earlier gen i7s and I've dropped 16GB of memory and 1TB spinning disks in each one. I've never been happier with a cluster. I've been free to spin whatever I want up and it just works. Using RKE to setup Kubernetes takes about 5 minutes to build a cluster and I'm using Longhorn for replicated container native storage. I've run experiments for work and my own experience. Whether its Redis / Cassandra cluster or a Kubernetes Operator I wrote, it can always handle it.
In my experience memory is a much bigger bottleneck than CPU, net IO or disk iops.
At the moment my favorite home lab cluster is a bunch of old laptops. They're quad cores too but I've dropped 16 GB of memory in each. It's doing a lot better. Right now the cluster is using 21GB of memory.
I'm able to run everything I've wanted as well as the occasional experiment to try something out.
If I could leave you with something - get more memory than you think you will need. That's been my experience anyway.
Consider checking the following: - Number of CPUs. The number of CPUs is not CPU cores but Kubernetes millicores. 1 vCPU on CR = 400m (4/10 of a CPU core). - Background processes - not recommended [https://stackoverflow.com/questions/61154349/google-cloud-ru...] - Container size on disk. Google Cloud is fast. However they need to cap the moving container images around as they can't use all the network capacity for moving a 400MB Docker image. If you're not already, use multi-stage builds with tiny final images.
Looks like Joe Rogan's advice was on the money
I remember a few people getting caught. What's funny about this is that the top players were still better than the AI running on a laptop (back then) and the cheaters occasionally got knocked out early on.
Your hosts are still packed with a bunch libraries and services (sshd for example) that should probably be updated with regularity.
I echo a lot of what you say here regarding run anywhere and not marrying some giant vendor.
In some cases they create obvious bots that make each side go "Aha! See their supporters are all bots!"
If the maker of goods is not in good standing for offsetting their impact then the importing state introduces taxes for that deficiency and has it handled in the importing country.
The second is the idea of showing the actual cost of offsetting. Having honest conversations about that will promote local and incentivize R&D to reduce that cost. Fantastic stuff.
---
I wonder how to take this? Is it good to use these as a recommended approach? I would bet that there could be good examples and bad? Is any of these the leader in terms of approach used for their domain? (Graph databases for example).
Could it be that these are `just interesting` but not authoritative?
We're all tired of the obligatory "I wonder how long before Google deprecates this...;)" but I wonder if Google would get better adoption if they explicitly published what would be needed for them not to drop the product... 1M monthly users? 3M? ...xM? or that when a project actually reaches that point.
Can't Facebook keep EU data in the EU and "deploy" ADs to the various regions?
I'm not sure I understand why the data has to go the the US. The only thing I can image is that AI models are better with larger data sets. But even then it's not too awful to have an EU ML model and another for the rest of the world. Is it enough to exit the whole EU market?
Apple is also putting up a very poor "training" program (Independent Repair Program) so that it looks like they're entertaining "right to repair" while they ensure the program is not adequate enough for people taking the course to handle even the most basic repairs.
Apple also locks it all up in tight NDAs so that they can't discuss this publicly.