Category Archives: Network

Internet. Networking.

HTTP/3 Explained

I’m happy to tell that the booklet HTTP/3 Explained is now ready for the world. It is entirely free and open and is available in several different formats to fit your reading habits. (It is not available on dead trees.)

The book describes what HTTP/3 and its underlying transport protocol QUIC are, why they exist, what features they have and how they work. The book is meant to be readable and understandable for most people with a rudimentary level of network knowledge or better.

These protocols are not done yet, there aren’t even any implementation of these protocols in the main browsers yet! The book will be updated and extended along the way when things change, implementations mature and the protocols settle.

If you find bugs, mistakes, something that needs to be explained better/deeper or otherwise want to help out with the contents, file a bug!

It was just a short while ago I mentioned the decision to change the name of the protocol to HTTP/3. That triggered me to refresh my document in progress and there are now over 8,000 words there to help.

The entire HTTP/3 Explained contents are available on github.

If you haven’t caught up with HTTP/2 quite yet, don’t worry. We have you covered for that as well, with the http2 explained book.

I’m leaving Mozilla

It’s been five great years, but now it is time for me to move on and try something else.

During these five years I’ve met and interacted with a large number of awesome people at Mozilla, lots of new friends! I got the chance to work from home and yet work with a global team on a widely used product, all done with open source. I have worked on internet protocols during work-hours (in addition to my regular spare-time working with them) and its been great! Heck, lots of the HTTP/2 development and the publication of that was made while I was employed by Mozilla and I fondly participated in that. I shall forever have this time ingrained in my memory as a very good period of my life.

I had already before I joined the Firefox development understood some of the challenges of making a browser in the modern era, but that understanding has now been properly enriched with lots of hands-on and code-digging in sometimes decades-old messy C++, a spaghetti armada of threads and the wild wild west of users on the Internet.

A very big thank you and a warm bye bye go to everyone of my friends at Mozilla. I won’t be far off and I’m sure I will have reasons to see many of you again.

My last day as officially employed by Mozilla is December 11 2018, but I plan to spend some of my remaining saved up vacation days before then so I’ll hand over most of my responsibilities way before.

The future is bright but unknown!

I don’t yet know what to do next.

I have some ideas and communications with friends and companies, but nothing is firmly decided yet. I will certainly entertain you with a totally separate post on this blog once I have that figured out! Don’t worry.

Will it affect curl or other open source I do?

I had worked on curl for a very long time already before joining Mozilla and I expect to keep doing curl and other open source things even going forward. I don’t think my choice of future employer should have to affect that negatively too much, except of course in periods.

With me leaving Mozilla, we’re also losing Mozilla as a primary sponsor of the curl project, since that was made up of them allowing me to spend some of my work days on curl and that’s now over.

Short-term at least, this move might increase my curl activities since I don’t have any new job yet and I need to fill my days with something…

What about toying with HTTP?

I was involved in the IETF HTTPbis working group for many years before I joined Mozilla (for over ten years now!) and I hope to be involved for many years still. I still have a lot of things I want to do with curl and to keep curl the champion of its class I need to stay on top of the game.

I will continue to follow and work with HTTP and other internet protocols very closely. After all curl remains the world’s most widely used HTTP client.

Can I enter the US now?

No. That’s unfortunately not related, and I’m not leaving Mozilla because of this problem and I unfortunately don’t expect my visa situation to change because of this change. My visa counter is now showing more than 214 days since I applied.

HTTP/3

The protocol that’s been called HTTP-over-QUIC for quite some time has now changed name and will officially become HTTP/3. This was triggered by this original suggestion by Mark Nottingham.

The QUIC Working Group in the IETF works on creating the QUIC transport protocol. QUIC is a TCP replacement done over UDP. Originally, QUIC was started as an effort by Google and then more of a “HTTP/2-encrypted-over-UDP” protocol.

When the work took off in the IETF to standardize the protocol, it was split up in two layers: the transport and the HTTP parts. The idea being that this transport protocol can be used to transfer other data too and its not just done explicitly for HTTP or HTTP-like protocols. But the name was still QUIC.

People in the community has referred to these different versions of the protocol using informal names such as iQUIC and gQUIC to separate the QUIC protocols from IETF and Google (since they differed quite a lot in the details). The protocol that sends HTTP over “iQUIC” was called “hq” (HTTP-over-QUIC) for a long time.

Mike Bishop scared the room at the QUIC working group meeting in IETF 103 when he presented this slide with what could be thought of almost a logo…

On November 7, 2018 Dmitri of Litespeed announced that they and Facebook had successfully done the first interop ever between two HTTP/3 implementations. Mike Bihop’s follow-up presentation in the HTTPbis session on the topic can be seen here. The consensus in the end of that meeting said the new name is HTTP/3!

No more confusion. HTTP/3 is the coming new HTTP version that uses QUIC for transport!

curl up 2019 will happen in Prague

The curl project is happy to invite you to the city of Prague, the Czech Republic, where curl up 2019 will take place.

curl up is our annual curl developers conference where we gather and talk Internet protocols, curl’s past, current situation and how to design its future. A weekend of curl.

Previous years we’ve gathered twenty-something people for an intimate meetup in a very friendly atmosphere. The way we like it!

In a spirit to move the meeting around to give different people easier travel, we have settled on the city of Prague for 2019, and we’ll be there March 29-31.

Sign up now!

Symposium on the Future of HTTP

This year, we’re starting off the Friday afternoon with a Symposium dedicated to “the future of HTTP” which is aimed to be less about curl and more about where HTTP is and where it will go next. Suitable for a slightly wider audience than just curl fans.

That’s Friday the 29th of March, 2019.

Program and talks

We are open for registrations and we would love to hear what you would like to come and present for us – on the topics of HTTP, of curl or related matters. I’m sure I will present something too, but it becomes a much better and more fun event if we distribute the talking as much as possible.

The final program for these days is not likely to get set until much later and rather close in time to the actual event.

The curl up 2019 wiki page is where you’ll find more specific details appear over time. Just go back there and see.

Helping out and planning?

If you want to follow the planning, help out, offer improvements or you have questions on any of this? Then join the curl-meet mailing list, which is dedicated for this!

Free or charge thanks to sponsors

We’re happy to call our event free, or “almost free” of charge and we can do this only due to the greatness and generosity of our awesome sponsors. This year we say thanks to Mullvad, Sticker Mule, Apiary and Charles University.

There’s still a chance for your company to help out too! Just get in touch.

curl up 2019 with logos

DNS-over-HTTPS is RFC 8484

The protocol we fondly know as DoH, DNS-over-HTTPS, is now  officially RFC 8484 with the official title “DNS Queries over HTTPS (DoH)”. It documents the protocol that is already in production and used by several client-side implementations, including Firefox, Chrome and curl. Put simply, DoH sends a regular RFC 1035 DNS packet over HTTPS instead of over plain UDP.

I’m happy to have contributed my little bits to this standard effort and I’m credited in the Acknowledgements section. I’ve also implemented DoH client-side several times now.

Firefox has done studies and tests in cooperation with a CDN provider (which has sometimes made people conflate Firefox’s DoH support with those studies and that operator). These studies have shown and proven that DoH is a working way for many users to do secure name resolves at a reasonable penalty cost. At least when using a fallback to the native resolver for the tricky situations. In general DoH resolves are slower than the native ones but in the tail end, the absolutely slowest name resolves got a lot better with the DoH option.

To me, DoH is partly necessary because the “DNS world” has failed to ship and deploy secure and safe name lookups to the masses and this is the one way applications “one layer up” can still secure our users.

DoH in curl

DNS-over-HTTPS (DoH) is being designed (it is not an RFC quite yet but very soon!) to allow internet clients to get increased privacy and security for their name resolves. I’ve previously explained the DNS-over-HTTPS functionality within Firefox that ships in Firefox 62 and I did a presentation about DoH and its future in curl at curl up 2018.

We are now introducing DoH support in curl. I hope this will not only allow users to start getting better privacy and security for their curl based internet transfers, but ideally this will also provide an additional debugging tool for DoH in other clients and servers.

Let’s take a look at how we plan to let applications enable this when using libcurl and how libcurl has to work with this internally to glue things together.

How do I make my libcurl transfer use DoH?

There’s a primary new option added, which is the “DoH URL”. An application sets the CURLOPT_DOH_URL for a transfer, and then libcurl will use that service for resolving host names. Easy peasy. There should be nothing else in the transfer that changes or appears differently. It’ll just resolve the host names over DoH instead of using the default resolver!

What about bootstrap, how does libcurl find the DoH server’s host name?

Since the DoH URL itself typically is given using a host name, that first host name will be resolved using the normal resolver – or if you so desire, you can provide the IP address for that host name with the CURLOPT_RESOLVE option just like you can for any host name.

If done using the resolver, the resolved address will then be kept in libcurl’s DNS cache for a short while and the DoH connection will be kept in the regular connection pool with the other connections, making subsequent DoH resolves on the same handle much faster.

How do I use this from the command line?

Tell curl which DoH URL to use with the new –doh-url command line option:

$ curl --doh-url https://dns-server.example.com https://www.example.com

How do I make my libcurl code use this?

curl = curl_easy_init();
curl_easy_setopt(curl, CURLOPT_URL,
                 "https://curl.haxx.se/");
curl_easy_setopt(curl, CURLOPT_DOH_URL,
                 "https://doh.example.com/");
res = curl_easy_perform(curl);

Internals

Internally, libcurl itself creates two new easy handles that it adds to the existing multi handles and they are then performing two HTTP requests while the original transfer sits in the “waiting for name resolve” state. Once the DoH requests are completed, the original transfer’s state can progress and continue on.

libcurl handles parallel transfers perfectly well already and by leveraging the already existing support for this, it was easy to add this new functionality and still work non-blocking and even event-based correctly depending on what libcurl API that is being used.

We had to add a new little special thing that makes libcurl handle the end of a transfer in a new way since there are now easy handles that are created and added to the multi handle entirely without the user’s knowledge, so the code also needs to remove and delete those handles when they’re done serving their purposes.

Was this hard to add to a 20 year old code base?

Actually, no. It was surprisingly easy, but then I’ve also worked on a few different client-side DoH implementations already so I had gotten myself a clear view of how I wanted the functionality to work plus the fact that I’m very familiar with the libcurl internals.

Plus, everything inside libcurl is already using non-blocking code and the multi interface paradigms so the foundation for adding parallel transfers like this was already in place.

The entire DoH patch for curl, including documentation and test cases, was a mere 1500 lines.

Ship?

This is merged into the master branch in git and is planned to ship as part of the next release: 7.62.0 at the end of October 2018.

much faster curl uploads on Windows with a single tiny commit

These days, operating system kernels provide TCP/IP stacks that can do really fast network transfers. It’s not even unusual for ordinary people to have gigabit connections at home and of course we want our applications to be able take advantage of them.

I don’t think many readers here will be surprised when I say that fulfilling this desire turns out much easier said than done in the Windows world.

Autotuning?

Since Windows 7 / 2008R2, Windows implements send buffer autotuning. Simply put, the faster transfer and longer RTT the connection has, the larger the buffer it uses (up to a max) so that more un-acked data can be outstanding and thus enable the system to saturate even really fast links.

Turns out this useful feature isn’t enabled when applications use non-blocking sockets. The send buffer isn’t increased at all then.

Internally, curl is using non-blocking sockets and most of the code is platform agnostic so it wouldn’t be practical to switch that off for a particular system. The code is pretty much independent of the target that will run it, and now with this latest find we have also started to understand why it doesn’t always perform as well on Windows as on other operating systems: the upload buffer (SO_SNDBUF) is fixed size and simply too small to perform well in a lot of cases

Applications can still enlarge the buffer, if they’re aware of this bottleneck, and get better performance without having to change libcurl, but I doubt a lot of them do. And really, libcurl should perform as good as it possibly can just by itself without any necessary tuning by the application authors.

Users testing this out

Daniel Jelinski brought a fix for this that repeatedly poll Windows during uploads to ask for a suitable send buffer size and then resizes it on the go if it deems a new size is better. In order to figure out that if this patch is indeed a good idea or if there’s a downside for some, we went wide and called out for users to help us.

The results were amazing. With speedups up to almost 7 times faster, exactly those newer Windows versions that supposedly have autotuning can obviously benefit substantially from this patch. The median test still performed more than twice as fast uploads with the patch. Pretty amazing really. And beyond weird that this crazy thing should be required to get ordinary sockets to perform properly on an updated operating system in 2018.

Windows XP isn’t affected at all by this fix, and we’ve seen tests running as VirtualBox guests in NAT-mode also not gain anything, but we believe that’s VirtualBox’s “fault” rather than Windows or the patch.

Landing

The commit is merged into curl’s master git branch and will be part of the pending curl 7.61.1 release, which is due to ship on September 5, 2018. I think it can serve as an interesting case study to see how long time it takes until Windows 10 users get their versions updated to this.

Table of test runs

The Windows versions, and the test times for the runs with the unmodified curl, the patched one, how much time the second run needed as a percentage of the first, a column with comments and last a comment showing the speedup multiple for that test.

Thank you everyone who helped us out by running these tests!

Version Time vanilla Time patched New time Comment speedup
6.0.6002 15.234 2.234 14.66% Vista SP2 6.82
6.1.7601 8.175 2.106 25.76% Windows 7 SP1 Enterprise 3.88
6.1.7601 10.109 2.621 25.93% Windows 7 Professional SP1 3.86
6.1.7601 8.125 2.203 27.11% 2008 R2 SP1 3.69
6.1.7601 8.562 2.375 27.74% 3.61
6.1.7601 9.657 2.684 27.79% 3.60
6.1.7601 11.263 3.432 30.47% Windows 2008R2 3.28
6.1.7601 5.288 1.654 31.28% 3.20
10.0.16299.309 4.281 1.484 34.66% Windows 10, 1709 2.88
10.0.17134.165 4.469 1.64 36.70% 2.73
10.0.16299.547 4.844 1.797 37.10% 2.70
10.0.14393 4.281 1.594 37.23% Windows 10, 1607 2.69
10.0.17134.165 4.547 1.703 37.45% 2.67
10.0.17134.165 4.875 1.891 38.79% 2.58
10.0.15063 4.578 1.907 41.66% 2.40
6.3.9600 4.718 2.031 43.05% Windows 8 (original) 2.32
10.0.17134.191 3.735 1.625 43.51% 2.30
10.0.17713.1002 6.062 2.656 43.81% 2.28
6.3.9600 2.921 1.297 44.40% Windows 2012R2 2.25
10.0.17134.112 5.125 2.282 44.53% 2.25
10.0.17134.191 5.593 2.719 48.61% 2.06
10.0.17134.165 5.734 2.797 48.78% run 1 2.05
10.0.14393 3.422 1.844 53.89% 1.86
10.0.17134.165 4.156 2.469 59.41% had to use the HTTPS endpoint 1.68
6.1.7601 7.082 4.945 69.82% over proxy 1.43
10.0.17134.165 5.765 4.25 73.72% run 2 1.36
5.1.2600 10.671 10.157 95.18% Windows XP Professional SP3 1.05
10.0.16299.547 1.469 1.422 96.80% in a VM runing on Linux 1.03
5.1.2600 11.297 11.046 97.78% XP 1.02
6.3.9600 5.312 5.219 98.25% 1.02
5.2.3790 5.031 5 99.38% Windows 2003 1.01
5.1.2600 7.703 7.656 99.39% XP SP3 1.01
10.0.17134.191 1.219 1.531 125.59% FTP 0.80
TOTAL 205.303 102.271 49.81% 2.01
MEDIAN 43.51% 2.30

quic wg interim Kista

The IETF QUIC working group had its fifth interim meeting the other day, this time in Kista, Sweden hosted by Ericsson. For me as a Stockholm resident, this was ridiculously convenient. Not entirely coincidentally, this was also the first quic interim I attended in person.

We were 30 something persons gathered in a room without windows, with another dozen or so participants joining from remote. This being a meeting in a series, most people already know each other from before so the atmosphere was relaxed and friendly. Lots of the participants have also been involved in other protocol developments and standards before. Many familiar faces.

Schedule

As QUIC is supposed to be done “soon”, the emphasis is now a lot to close issues, postpone some stuff to “QUICv2” and make sure to get decisions on outstanding question marks.

Kazuho did a quick run-through with some info from the interop days prior to the meeting.

After MT’s initial explanation of where we’re at for the upcoming draft-13, Ian took us a on a deep dive into the Stream 0 Design Team report. This is a pretty radical change of how the wire format of the quic protocol, and how the TLS is being handled.

The existing draft-12 approach…

Is suggested to instead become…

What’s perhaps the most interesting take away here is that the new format doesn’t use TLS records anymore – but simplifies a lot of other things. Not using TLS records but still doing TLS means that a QUIC implementation needs to get data from the TLS layer using APIs that existing TLS libraries don’t typically provide. PicoTLS, Minq, BoringSSL. NSS already have or will soon provide the necessary APIs. Slightly behind, OpenSSL should offer it in a nightly build soon but the impression is that it is still a bit away from an actual OpenSSL release.

EKR continued the theme. He talked about the quic handshake flow and among other things explained how 0-RTT and early data works. Taken from that context, I consider this slide (shown below) fairly funny because it makes it look far from simple to me. But it shows communication in different layers, and how the acks go, etc.

HTTP

Mike then presented the state of HTTP over quic. The frames are no longer that similar to the HTTP/2 versions. Work is done to ensure that the HTTP layer doesn’t need to refer or “grab” stream IDs from the transport layer.

There was a rather lengthy discussion around how to handle “placeholder streams” like the ones Firefox uses over HTTP/2 to create “anchors” on which to make dependencies but are never actually used over the wire. The nature of the quic transport makes those impractical and we talked about what alternatives there are that could still offer similar functionality.

The subject of priorities and dependencies and if the relative complexity of the h2 model should be replaced by something simpler came up (again) but was ultimately pushed aside.

QPACK

Alan presented the state of QPACK, the HTTP header compression algorithm for hq (HTTP over QUIC). It is not wire compatible with HPACK anymore and there have been some recent improvements and clarifications done.

Alan also did a great step-by-step walk-through how QPACK works with adding headers to the dynamic table and how it works with its indices etc. It was very clarifying I thought.

The discussion about the static table for the compression basically ended with us agreeing that we should just agree on a fairly small fixed table without a way to negotiate the table. Mark said he’d try to get some updated header data from some server deployments to get another data set than just the one from WPT (which is from a single browser).

Interop-testing of QPACK implementations can be done by encode  + shuffle + decode a HAR file and compare the results with the source data. Just do it – and talk to Alan!

And the first day was over. A fully packed day.

ECN

Magnus started off with some heavy stuff talking Explicit Congestion Notification in QUIC and it how it is intended to work and some remaining issues.

He also got into the subject of ACK frequency and how the current model isn’t ideal in every situation, causing to work like this image below (from Magnus’ slide set):

Interestingly, it turned out that several of the implementers already basically had implemented Magnus’ proposal of changing the max delay to min(RTT/4, 25 ms) independently of each other!

mvfst deployment

Subodh took us on a journey with some great insights from Facebook’s deployment of mvfast internally, their QUIC implementation. Getting some real-life feedback is useful and with over 100 billion requests/day, it seems they did give this a good run.

Since their usage and stack for this is a bit use case specific I’m not sure how relevant or universal their performance numbers are. They showed roughly the same CPU and memory use, with a 70% RPS rate compared to h2 over TLS 1.2.

He also entertained us with some “fun issues” from bugs and debugging sessions they’ve done and learned from. Awesome.

The story highlights the need for more tooling around QUIC to help developers and deployers.

Load balancers

Martin talked about load balancers and servers, and how they could or should communicate to work correctly with routing and connection IDs.

The room didn’t seem overly thrilled about this work and mostly offered other ways to achieve the same results.

Implicit Open

During the last session for the day and the entire meeting, was mt going through a few things that still needed discussion or closure. On stateless reset and the rather big bike shed issue: implicit open. The later being the question if opening a stream with ID N + 1 implicitly also opens the stream with ID N. I believe we ended with a slight preference to the implicit approach and this will be taken to the list for a consensus call.

Frame type extensibility

How should the QUIC protocol allow extensibility? The oldest still open issue in the project can be solved or satisfied in numerous different ways and the discussion waved back and forth for a while, debating various approaches merits and downsides until the group more or less agreed on a fairly simple and straight forward approach where the extensions will announce support for a feature which then may or may involve one or more new frame types (to be in a registry).

We proceeded to discuss other issues all until “closing time”, which was set to be 16:00 today. This was just two days of pushing forward but still it felt quite intense and my personal impression is that there were a lot of good progress made here that took the protocol a good step forward.

The facilities were lovely and Ericsson was a great host for us. The Thursday afternoon cakes were great! Thank you!

Coming up

There’s an IETF meeting in Montreal in July and there’s a planned next QUIC interim probably in New York in September.

curl, http2 and quic on the Changelog

Three years ago I talked on a changelog episode about curl just having turned 17 years old and what it has meant for me etc.

Fast forward three years, 146 changelog episodes later and now curl has turned 20 years and I was again invited and joined the lovely hosts of the changelog podcast, Adam and Jerod.

Changelog episode 299

We talked curl of course but we also spent time talking about where HTTP/2 is and how QUIC is coming around and a little about why and how its UDP nature makes things a little different. If you’re into either curl or web transport, I hope you’ll find it interesting.

Now at 1000 mbit

A little over six years since I got the fiber connection installed to my house. Back then, on a direct question to my provider, they could only offer 100/100 mbit/sec so that’s what I went with. Using my Telia Öppen Fiber and Tyfon (subsequently bought by Bahnhof) as internet provider.

In the spring of 2017 I bumped the speed to 250/100 mbit/sec to see if I would notice and actually take advantage of the extra speed. Lo and behold, I actually feel and experience the difference – frequently. When I upgrade my Linux machines or download larger images over the Internet, I frequently do that at higher speeds than 10MB/sec now and thus my higher speed saves me time and offers improved convenience.

However, “Öppen Fiber” is a relatively expensive provider for little gain for me. The “openness” that allows me to switch between providers isn’t really something that gives much benefit once you’ve picked a provider you like, it’s then mostly a way for a middle man to get an extra cut. 250mbit/sec from Bahnhof cost me 459 SEK/month (55 USD) there.

Switching to Bahnhof to handle both the fiber and the Internet connection is a much better deal for me, price wise. I get an upgraded connection to a 1000/1000 mbit/sec for a lower monthly fee. I’ll now end up paying 399/month (48 USD)  (299 SEK/month the first 24 months). So slightly cheaper for much more speed!

My household typically consists of the following devices that are used for accessing the web regularly:

  • 4 smart phones
  • 1 iPad
  • 4 laptops
  • 3 desktop computers
  • 1 TV computer

Our family of 4 consumes around 120GB average weeks. Out of this, Youtube is the single biggest hogger with almost 30% of our total bandwidth. I suppose this says something about the habits of my kids…

Out of these 13 most frequently used devices in our local network only 5 are RJ45-connected, the rest are WiFi.

Switch-over

I was told the switch-over day was May 15th, and at 08:28 in the morning my existing connection went away. I took that as the start signal. I had already gotten a box from Bahnhof with the new media converter to use.

I went downstairs and started off my taking a photo of the existing installation…

So I unscrewed that old big thing from the wall and now my installation instead looks like

You can also see the Ethernet cable already jacked in.

Once connected, I got a link at once and then I spent another few minutes to try to “register” with my user name and password until I figured out that my router has 1.1.1.1 hardcoded as DNS server and once I cleared that, the login-thing worked as it should and I could tell Bahnhof that I’m a legitimate user and woof, my mosh session magically reconnected again etc.

All in all, I was offline for shorter than 30 minutes.

Speeds and round-trips

These days a short round-trip is all the rage and is often more important than high bandwidth when browsing the web. I’m apparently pretty close to the Stockholm hub for many major services and I was a bit curious how my new operator would compare.

To my amazement, it’s notably faster. google.com went from 2.3ms to 1.3ms ping time, 1.1.1.1 is at 1.3ms, facebook.com is 1.0ms away.  My own server is 1.2ms away and amusingly even if I’m this close to the main server hosting the curl web site, the fastly CDN still outperforms it so curl.haxx.se is an average 1.0ms from me.

So, the ping times were notably reduced. The bandwidth is truly at gigabit speeds in both directions according to bredbandskollen.se, which is probably the most suitable speed check site in Sweden.

A rather smooth change so far. Let’s hope it stays this way.