8 ms·
Some notes on high speed networking on PCs
- walterbell 8y ago> the kernel community is working on AF_XDP See recent Intel presentation on Time Sensitive Networking, https://schd.ws/hosted_files/elciotna18/b6/ELC-2018-USA-TSNonLinux.pdf https://schd.ws/hosted_files/elciotna18/b6/ELC-2018-USA-TSNo...
- lossolo 8y agoTSN is interesting for special use cases like the one they presented in the presentation. I've impression that those two are solving two different problems. TSN focus on deterministic, precise timing of packets and low latency while sacrificing throughput. AF_XDP focuses mainly on throughput.
- nickpsecurity 8y agoBack when I looked into high-speed networking in small spaces, my idea was doing it over PCI directly bypassing Ethernet. I really wanted to do something like SGI's NUMAlink but PCI be cheaper with commodity parts available. You'd need a switch for it, though. I ended up finding a company selling them. So, there's always that kind of thing to consider. I wonder if 10-100Gbps have better price/performance than TCP/IP over PCI by now. I cant even rember the company or tech's name.
- wmf 8y agoThat's possible with PLX/Broadcom ExpressFabric but those switches have fallen far behind Ethernet. The fastest PCIe switch is ~768 Gbps while ~6500 Gbps Ethernet switches are available.
- namibj 8y agoYou need to consider that these PCIe switches you talk about are (afaik, at least there are big ones) single-die, compared to the probably even multi-board construction of the Ethernet switches. They do support and are in at least one document advertised to be used in a switching fabric, where you create a mesh between the nodes you try to connect. Basically, instead of plugging every computer into the same PCIe switch, you fill the PCIe switches half full with computers, maybe even a little less/add some that don't connect directly to computers, and add some interconnections between the switches. The mathematics that describe how one would want to arrange the small switches date back at least to when automated telephone switches became common. A similar version, though with separated inputs and outputs, is known as a Crossbar Switch. Did you by chance come across something smaller than a full-blown, PCIe3x4 per node, TOR switch? I don't need that much capacity, but a cheap enough high speed switch would allow using some slower (individually) servers that got kicked out from the previous owner, instead of getting new, larger machines. A certain level of microservice-thinking/scale of (distributed) Haskell/Erlang does often require much faster connections than one would like to afford with Ethernet, to keep the coding overhead for the sharding to a minimum. At a certain point one can stop worrying about minimizing transfers, and just rely on page-faulting and similar things for less-critical (in a performance sense) data accesses.
- wmf 8y agoI was talking about single-chip switches. The thing about Clos is that it has non-linear scaling, so if you use a chip with lower capacity you end up needing far more chips. All of these chips have smaller sizes available.
- rasz 8y agoThere was a product in late nineties that did P2P links over ATA66/100 IDE interface. No drivers needed, just write/read to a virtual filesystem on simulated drive.
- daveguy 8y agoI wanted to do the same thing with PCI!! I thought the bandwidth was large enough that it should be a great direct interface. Unfortunately synchronizing between two pci masters was going to be tricky and the high frequency development required for PCIe requires some pricey diagnostic equipment. I ran across those off the shelf switches / interfaces, but they were a lot more expensive that 10gb ethernet at the time and they weren't even available for high lane count pcie 2 and pcie 3. I think now that 10gb Ethernet prices are coming down to about $100 per card and switch port there's no way specialized direct pcie would compete off the shelf. Still I think it would be cool to have a direct 32 lane pcie 3 of 4. It could operate as fast as ram speeds. If you connected two 2P beasts it would be close to having a 4P. Of course there would probably be protocol issues and overhead as the no-two-masters issue would mean you can't use the native protocol directly.
- nickpsecurity 8y agoI didnt think about the specialized, diagnostic gear. That couldve priced me out. Far as your other idea, a Distributed, Shared Memory setup was one of my potential applications. I was looking at Beowulf clusters and old MPP systems trying to find ways to get more CPU's with large RAM and simplified programming. I found quite a few software implementations of DSM model. Infiniband and Myrinet were main thing in networking. So, PCI idea was partly about connecting them with something that already had DMA. That at least led me to NUMAscale's products for scaling AMD systems.
- exikyut 8y agoNote this comment in this subthread: https://news.ycombinator.com/item?id=17276861 https://news.ycombinator.com/item?id=17276861
- namibj 8y agoDon't worry, the chip you are searching is branded ExpressFabric and sold/made by Broadcom. It's AFAIK only PCIe4, and you get ones with up to iirc 96 lanes per chip. They support to be connected into a mesh, even adding shared PCIe devices like network cards, GPUs, NVMe and SAS/SATA HBAs is possible, though you might have difficulties with the GPUs in particular. The others, if you are careful when buying, should do well if they can handle SR-IOV, as (from their view), the mesh switch they are attached to is the root, and the computers attached to the mesh take the place a VM would normally occupy. If you could get the smallest chip they make, it should not be hard to get a PCB made for it that fans out to e.g. a set of USB-C receptacles. One such receptacle can handle 2 PCIe3 lanes, and some power/USB-2 on the side. The main reason being the relatively low price, compared to e.g. mini SAS HD, which has a similar density and twice the lanes per connector, but also a slightly higher target impedance at the upper end of the PCIe spec, whereas USB-C is specifically targeting the perfect PCIe impedance.
- voltagex_ 8y agoDo you think we'll ever see 10G networking "just work" at home? I mean, I can plug in a couple of PCs into a relatively cheap switch with relatively cheap cabling and get somewhere approximating 1 gigabit right now.
- _jal 8y agoI know what you mean, but it has gotten a lot cheaper. There are sub-$1K 10G switches and consumer-ish onboard-10G motherboards. Probably, something popular needs building that demands 10G-in-the-home to use, in order to get volume up so pricing can come down a bit. VR, if it actually does hit this time, is a candidate. As far as fragility, copper is the way to go at home, unless you're entirely pet and child-free or have your own cage in the garage.
- JohnJamesRambo 8y agoCan you expand on that last sentence? I don’t understand it but it sounds interesting.
- Firerouge 8y agoFiber optic cables are too fragile for home use. If bent to any extreme the cable will be ruined. Hence only useful in homes with protected installations like a server enclosures.
- NightlyDev 8y agoThat's a myth. Fiberoptic cables are actually really robust.
- StillBored 8y agoWell POF is, real glass still has bend diameter limitations, but most people aren't using single mode longer fiber runs that need glass. So, yes, all the common stuff is pretty flexible, because its plastic.
- Spooky23 8y ago10gb on VDI is fun when your storage can absorb line rate spikes. Things break in new ways and are fun to troubleshoot.
- exikyut 8y agoThis sounds kind of interesting. What happened?
- Spooky23 8y agoThe constraints moved! We had a write intensive workload for a user community that would generate short (~5s) periods of high traffic (6-9Gb) against a file server — this would impact everyone because sessions would be interrupted. It was devilishly difficult to troubleshoot because it was lost in the statistics, aggregated in 30-100s chunks. Fixing that revealed another constraint where we hit a constraint with the file server process itself. It’s one of these things that was interesting because the storage with fancy SSD can handle the workload no sweat. In the old days, monitoring would catch the io constraint first.
- exikyut 8y agoHmm. This highlights the importance of statistics AND insane threshold edge detection. Good one to file away, thanks.
- Spooky23 8y agoYeah, we’re working on a way to grab real-time stats during peak load conditions to help correct for outliers lost in averages. Another key thing is to be aware where heavy write load generators are on SMB, and scope the SMB shares around them to scope failure domains. There are limits, and the vendors (Microsoft or 3rd party NAS) have a hard time finding them too.
- wemdyjreichert 8y agoWell, that was... bellicose. No company is your friend, but some are still better than others. It is true that companies work in their own self-interests, but those interests may differ. Apple has done quite well without selling user data, and has cultivate a brand that makes money doing something else. Facebook has not. The enemy of my enemy is my friend, even if they work in their own self-interest. It is in the self-interest of apple to keep customers and they might lose more than they gain selling user info. Again, no company is perfect. "Applying constant pressure" isn't perfect either, as governments are at least as bad as any company - their motivation is to stay in office. Apple is at least a better company then, say, Facebook.