Is there any advantage of running k3s if you want to keep etcd? I understand that most k3s performance gains come from etcd being replaced by sqlite but if you still want a HA control plane, sqlite won't cut it.
We've been using k3s' embedded etcd for as long as its existed, and it's great. Setting up the etcd cluster is dramatically simplified; let the first node generate a token and feed it to all the other nodes. Tons of other advantages to k3s; the single-binary deploy process, the built-in networking stack (which you can secure with Wireguard out-of-the-box), built-in ingress controller if you want one.
you can easily still use external etcd if you really need to.
But in general k3s can be HA without issue and scaled just as well as vanilla k8s.
The main advantage of it is that everything comes neatly packed into a single binary whereas the alternative would mean to have a multitude of services running for cluster provisioning.
Kubernetes in the end is basically an API server with multiple componets and k3s puts a nice bow around all of them.
There is nothing wrong with ONNX, but rather the limitations of PyTorch.
1. PyTorch model files are neither portable nor self-contained, because PyTorch model files are pickled Python classes containing weights so you need Python class code to run it.
Because it needs real Python code to run the models, PyTorch suffers from a numerous issues porting to other non-Python platforms such as ONNX.
PyTorch offers a way to export to ONNX but you will encounter various errors. [1]
Sure, you might be lucky enough to troubleshoot a specific model to export it to ONNX, but if your objective is to export arbitrary 964 models from a model zoo (TIMM) it is almost impossible.
2. There are organizational or cultural problems with it. Because of the above problem, PyTorch model needs to be designed with portability in mind from beginning. But porting & serving models are what engineers do, whereas researchers, who design models, don't care about it when writing papers. So it is often hard to use SOTA models that comes from an academic research.
> PyTorch offers a way to export to ONNX but you will encounter various errors. [1]
I mean sure, there are limitations, but this is greatly exaggerating their impact in my experience. I'd be curious to hear from anyone where these have been serious blockers, I've been exporting PyTorch models to ONNX (for CV applications) for the last couple of years without any major issues (and any issues that did pop up were resolved in a matter of hours).
How did this happen? a pickle is not a sensible storage format. it's insecure, hard to version, not very portable. isnt a model basically a big matrix of numbers?
Not in PyTorch. A model is Python dictionaries containing states and Python module/class objects. I don't know why the PyTorch team did this but that happened. Maybe it boils down to the point #2 I said.
This kind of thinking is exactly what's killing signal. Why would you want to gatekeep the security of others instead of making t as accessible as possible?
To me this feels like signal not understanding that their intended userbase and their actual userbase are very different, as I can't imagine the number of people that use signal solely for it's e2ee is comparable to the number of people that use it as their sms app.
Well, isn't the point that more people using signal strengthens the privacy/security of everyone? I would be sad if I wanted to send a message to gerty, and I find out he stopped using signal because of this.
It's not like I wouldn't still want to message him, right?
Convenience is extremely powerful in getting the layman to adopt this kind of tech, and I feel like it should be prioritized.
It's not built-in, of course, but if you can live with the quirky API, LMDB as an embedded key-value store will be hard to beat in terms of speed. It can also do "one-writer, many-readers" and transactions.
I wouldn't count on it unless you add some extra repo. RHEL releases are historically behind when it comes to software, which is claimed to provide more stability, but often older kernels in RHEL simply will not support the new underlying functionality. I am hopeful that if Podman ever has to make the decision to adapt at a modern pace with new software releases, or constantly try to backport fixes and code for old software, that they'll choose the modern approach.
RHEL 8.6 upgraded podman from v3.3 to v4.0.2, so v4.2 is not outside the realm of possibility in the next release. Since podman development is driven by Red Hat, I feel that podman upgrades get quite some leeway... Very impatient to test out the play kube functionality managed by systemd.
First, some software is updated frequently in RHEL, including podman. RHEL 8 was released in 2019 and it has podman 4.0.2 from earlier this year.
Second, even software that isn't updated often, or at all, in the base system might have newer releases available as modules. For example there are recent versions of Python in RHEL 8, with just the basic runtime so that you can use pip to install more packages.
Third, the RHEL kernel is updated much more than the corresponding LTS releases. The RHEL 8 kernel is closer to 5.15-ish than to the nominal 4.18 release from which it was forked. Pace for backports has slowed down a bit, but there's interest in keeping the kernel up to date because people are running RHEL 9 containers on the RHEL 8 kernel.
Given the module streams functionality in the package system in RHEL 8+, it's fairly plausible that it would be introduced using that path.
(Currently, there are 5 'container-tools' module streams listed, with podman versions including 1.0, 1.6, 3.0, 4.0 in stable streams, and 4.0.2 in the rolling stream.)
That is fair. I don't use Red Hat anymore really and their marketing and naming around modules was just confusing at the time. They took a common component of package management called repositories and made it confusing.
It's not just a contract though. France actually has a maritime border with Australia and its own interests in the Pacific -> the entire French Pacific strategy was wiped out by 3 "allies" behind its back.
There are basically three things that France can now do:
a) abandon New Caledonia entirely. This may happen anyway, as an independence referendum is scheduled for December and the balance is uncertain;
b) join AUKUS as a junior partner, very painful but logical;
c) embrace China out of spite or as a result of different geopolitical calculus. That would probably have significant realignment echos in Europe, because France has a big voice in the EU. I am a little afraid of this scenario, though I consider it unlikely to happen.
Edit: I remembered that Tahiti is at stake too. Not certain about local independence movement status.
It's more that Caledonia could chose to leave. But that only marginaly change the issue. France has other territories in the Pacific.
> b) join AUKUS as a junior partner, very painful but logical;
That would be very illogical. France interests are not aligned with the USA. Also they are a nuclear country with a permanent seat on the UN security council. They will never be a junior partner to an alliance with the USA or Australia let alone the UK.
> c) embrace China out of spite or as a result of different geopolitical calculus. That would probably have significant realignment echos in Europe
There is no need for spite or realignment. China is a diplomatic and commercial partner of the EU. The systemic opposition is strictly the doctrine of the USA. There is plenty to gain through the diplomatic route as long as you see China as an equal partner and not an inferior country. Lobbying China is harder than it used to be but not impossible.
There are plenty of alternative to these three scenarios anyway. The logical next steps would just be pushing for further for military integration in the EU (Afghanistan already left a sour taste in everyone's mouth anyway) and lobby to get the EU - Australia free trade agreement severely limited then carry on as usual. The relationship with the UK will see no progress as long as Johnson is PM so it seems useless to waste time on it.
> There is plenty to gain through the diplomatic route as long as you see China as an equal partner and not an inferior country.
Does China see other countries as equal partners? Their Howling Wolf diplomacy seems to kick in whenever anything they dislike appears in any media. See their current hatefest against Lithuania, a country with population smaller than Shanghai.
I for one am not very ready to sacrifice, e.g. freedom to criticize the Chinese system or to say that Taiwan is Taiwan and not Chinese Taipei or whatever. And I do not trust China commercially either, given their history of copying everything and then flooding markets with cheap knock-offs to undermine the original producers.
We have a load of our own problems here in the West, but we can at least discuss them mostly freely and the authoritarian developments (governmental and distributed alike) tend to get some pushback and dissent, and the dissidents do not end up in jail or shot. These are not the values of Xi's China and any partnership with it will end in Beijing dictating their demands.
> Does China see other countries as equal partners? Their Howling Wolf diplomacy seems to kick in whenever anything they dislike appears in any media.
China is an authoritarian country with a strategic vision which put it at odd with its closest neighbors. It's in no way an easy partner. Then again, neither is it an unreadable one nor a completely unreasonable one. What China wants is pretty clear: control of the South China sea and being the dominant soft power of the region. With this framework in mind, a lot can be achieved through negociations.
I am not saying countries should always be soft with China. Their track record record regarding human rights is worrying. I am very much in favor of using tit for tat measures regarding their distortion of competitive markets and their threats on the Chinese diaspora in Europe and the USA have to stop.
Still I am not convinced that a policy of systematic opposition and war mongering will do much good. Xi Jinping is already 68. His rule won't last eternally. It seems better to me to adopt a position of observation, containment and soft power influence than direct confrontation. I understand that Taiwan might be a victim of it which sadden me a bit but is something I am fine with from a realpolitik point of view. I consider this position better for Europe than following the USA in its opposition which I view as essentially motivated by a desire to protect the position they consider to be theirs in the international order rather than some kind of moral imperative.
"Xi Jinping is already 68. His rule won't last eternally."
Eternally not, but the recent progresses in longevity field might give him 20-30 more years at the helm.
This is an interesting strategic question that isn't discussed as much as it ought to. Even Western nations are evolving into gerontocracies. We should probably introduce some hard term limits. Already there are U.S. Senators that served more than 40 years.
> See their current hatefest against Lithuania, a country with population smaller than Shanghai.
The US isn't much better. They have a history of invading of destabilising the governments of countries they don't like.
> we can at least discuss them mostly freely and the authoritarian developments (governmental and distributed alike) tend to get some pushback and dissent, and the dissidents do not end up in jail or shot.
Tell that to all the countries who tried to implement communist systems of government... and found their leaders shot or in jail.
> I do not trust China commercially either, given their history of copying everything and then flooding markets with cheap knock-offs to undermine the original producers.
And I don't trust the US commercially, with their history of enforcing tariff-free markets on countries and flooding them with cheap US-made goods that wipe out domestic production leaving the countries in very dire straits (see e.g. https://en.wikipedia.org/wiki/Rice_production_in_Haiti).
This is where different countries have different experiences.
If I were a Latin American, I would be distrustful towards American power projection. But I am a Central European and our history with the U.S. is a lot better.
I am not trusting them blindly, but the worst import we had from the U.S. so far was a shallow-ish pop culture, plus some of the racial nonsense that really rubs the wrong way in a different context.
As a citizen of the UK, I'd argue that Reagan-style economics (which caused us to privatise a lot of our industry and public services in a such a way that it was effectively just a huge cash giveaway to private companies) and ridiculous copyright terms are our worst imports. You may be less effected by those kind of things in Central Europe though, and you could certainly argue, and you could certainly argue that these were our own stupidity rather than being enforced upon us.
> The US isn't much better. They have a history of invading of destabilising the governments of countries they don't like.
This is whataboutism and has nothing to do with discussion on France/EU's stance of China.
Reading on you comments, are you living in the past? We are discussing current politics.
> Tell that to all the countries who tried to implement communist systems of government... and found their leaders shot or in jail.
That was happening 40 years ago in current EU countries, whats your point?!
> with their history of enforcing tariff-free markets on countries and flooding them with cheap US-made goods that wipe out domestic production leaving the countries in very dire straits
Again derailing conversation from China to USA
Edit: go ahead and keep downvoting me for literally pointing out logical fallacies
US largely outsourced its manufacturing, so EU is not as dependant on US as other way around - EU to USA import export graph [0]
Treating EU like a tiny country that needs to rely on imports of finished products is discounting the fact EU is the largest trading block in the world.
Besides EU can decide to incentivise the manufacturing of key goods (like germany getting into electronics manufacture).
Also US cannot bully EU like it does other weaker countries.
> lobby to get the EU - Australia free trade agreement severely limited then carry on as usual
This type of response always baffles me. Australia here is not buying what France is selling. If they object to that, making it harder for Australians to buy other things France is selling is the opposite of clever.
And in reverse - Australia sells food and raw materials. Under what circumstances is it bad for France to trade bits of paper for actually useful physical goods?
Restricting trade is cutting of the nose to spite the face. Australia sells commodities, we don't care if any individual block trades with us. China tried clamping down on trade with Australia and it didn't achieve much. They've got to buy from someone else, that someone else's customers becomes Australia's customers.
> Australia here is not buying what France is selling. If they object to that, making it harder for Australians to buy other things France is selling is the opposite of clever.
Australia have been extremely untransparent regarding their position and how they conducted the whole negociation. They lied to their French counterparts during the summer and announced their withdrawal in an extermely clumsy and frankly disrespectful way. Why would you want to put in place treaties easing controls and trades with countries which have already proven they are dishonest?
> Restricting trade is cutting of the nose to spite the face.
No one talked about restrictions. I just expect France to lobby for things to stay as they are rather than go in the direction of easier trading.
What plausible risk is there that Australians won't honour the deals they make under a free trade deal? If someone buys a million dollars worth of Australian beans they're going to get tonnes of beans.
I don't fault the French if they don't see a benefit and want to be annoying in some sort of tit-for-tat strategy; but making trade harder is not going to make their lives better. People only trade when the trade improves their position.
A $60+ billion military deal is never going to be handled under a free trade agreement. I haven't read into the details but I doubt there was even any fraud involved; just standard political backstabbing. The French weren't given an opportunity but they are not worse off.
In France, fierce protectionism is part of the governmental policy longer than the Republic exists. This has seeped into EU policies as well, though not to the same extent.
> b) join AUKUS as a junior partner, very painful but logical;
Why is this logical? These partners have already shown that they are going to betray France's trust and lie to them. It would be better to be on your own than have unreliable partners, or worse ones that work against you.
Unreliable partners are Europe's mode of operation since Antiquity. The positive side is that European countries mostly know what to expect from one another. Xi's China is a new player whose intents are much less legible.
> These partners have already shown that they are going to betray France's trust and lie to them.
What? Betray France? So France jeopardizes a nation's national defense by failing to deliver critical defense assets, and this is spun into being a betrayal on France?
Are we supposed to put our blinders on to think that this issue is about France's commercial interests and not Australia's national security?
They hit their September deadline with the project, and all of the requested changes. The assets were on track, with the requested changes.
The betrayal issue comes with the very recent statements from Australia about supporting the project, whilst behind the scenes unbeknownst to France planning to scrap it.
To be fair, France has consistently betrayed its allies, throughout history. The start of WWII is worth looking at, where both France and Britain broke mutual defense treaty after mutual defense treaty appeasing Germany.
Culture runs deep. France adores Vercingetorix, who surrendered to Julius Caesar after "releasing" all the women and children to die of starvation between the Roman and French lines at the Battle Of Alesia.
Since WWII, France has been a fickle ally to the US at best.
The EU and (especially France and Germany) is not THAT critical of China as the US (and other countries...), and France has a big sense of national pride and no problems acting on it.
Is this in any way "good"? I don't know, but realisticaly i CAN see the EU siding more with China than the US in the long run.
Well, the big thing holding up a trade deal with China right now is that China decided to sanction several MEPs and the
European Parliament's human rights subcommittee for scrutinizing their forced labour camps in Xinjiang, which the European Parliament of course considered an attack on democracy itself. So if France used its sway to get the EU to align with China, that'd basically demonstrate that the French government's feelings had more power than the pretence of EU democracy.
Now... i would think it would go on the line of "change through rapprochement", and even this would not be THAT far fetched. China may be a dictatorial regime, but its an dictatorial regime that values good trade deals and making money. So getting concessions on not genociding the Uigurs (at least not in a way the world can be witness) are becoming more easy.
There are also alternative ways of sending a message, e.g. targeting of US soft power. While INSTEX was not a huge success in its time, an alternative payment mechanism that works around US sanctions aligns well with French (and even EU) interests and strongly against US.
Offering closer partnership to India, Indonesia, Malaysia and Singapore might be a viable alternate way forward for France - and they might find it in their interest too. More "indo" than "pacific" though.
They've already been working on this with India [1]. They have been working on building partnerships in the region for years, with a focus on India and Australia.
This is why they are deeply pissed off beyond the loss of this submarines deal. Basically the US have 'torpedoed' a big part of their long term strategy here.
India is naturally more inclined to cooperate with the UK, though. Colonial legacies are complicated and Indian elite tends to go study in English-speaking countries, so there is a lot of personal contact. It is much rarer to meet Indian students in continental EU.
There's a new trending joke in defense circles about how the opportunity was missed for a FAUKUS alliance, which would have been comically delicious to pronounce.
d) stand its ground independently, invest in its military to have the capability it needs, work with AUKUS if and when it is in its interests to do so. This includes taking a more 'non-aligned' stance on China, but that'd not be historically new (France was the first Western country to recognise the PRC and De Gaulle even advised Kennedy to do it sooner than later).
French vital interests are in Africa, not in Pacific. And the strategic situation in Sahel isn't good. Given that military budgets are always strained, it makes sense to concentrate on the core challenges and leaving the Pacific be.
France has strategic interests in Africa. And it has also vital interests in the Pacific because it actually has territories in the region.
France has the largest maritime EEZ in the world, owning 8% of the global EEZ areas. A lot of that is due to its territories in the Pacific and Indian Ocean. New Caledonia also has 25% of the world's nickel reserve. Defending your national territory is a vital interest.
Saying that it "should leave the Pacific be" is exactly what the US and Australia want (hence France being excluded from AUKUS), and exactly what it should not do.
You got me. I was only thinking about the territories themselves (arguably not crucially important) and forgot about the maritime EEZs that come with them. These seem to be rather important.
(Though probably not "vital" in the same sense as suppression of jihadism is. A massive growth of jihadism in North Africa could destabilize peace in France itself, while loss of Pacific territories would only be painful economically.)
Lets not pretend this was a mighty backstabbing by everyone else who france treated with nothing but kindness. I can't speak much for the US or Aus, but I wouldn't call them on good terms with the UK, beyond even the historical jibes.
I'm sure this is bad for France, don't get me wrong, but in terms of how their allies acted, its not particularly shocking or even abnormal.
It is, tho. Diplomacy is a sensible thing. It's not as much about the contract lost, but the way it was handled.
The contract was also a key piece for a deeper collaboration in the pacific area, one of the most important geopolitical sectors for the next 20 years.
France shared highly confidential material for months, while Australia already had decided to breach the contract. In international relationships among allies where trust is crucial, it's pretty big.
As I said, France has not been a strong ally of the UK (again, cant speak for US or Aus, but may be a similar situation there) in the past or even very recently, why you would expect differently the other way around I don't know.
Also doesn't look like this was a "breach" of contract, just a cancellation. A breach implies what was done is illegal in international law, a cancellation implies they simply went elsewhere.
