Re: [RFC] Add support for changing the flow ID of TCP connections
On 07/09/14 18:31, Navdeep Parhar wrote: On Wed, Jul 09, 2014 at 04:36:53PM +0200, Hans Petter Selasky wrote: On 07/08/14 21:17, Navdeep Parhar wrote: ... I think we need to design this to be as generic as possible. I have quite a bit of code that does this stuff but I haven't pushed it upstream or even offered it for review (yet). Hi, When will the non hardware related patches be available for review? I understand there are multiple ways to reach the same goal, and I think it would be great if we could agree on a common API for applications. Here is the kernel side of the patch: http://people.freebsd.org/~np/flow_pacing_kern.diff The registration parameters and the throttling parameters are probably cxgbe-centric, because that's what it was written for. We'll need to tidy up those structs certainly. And I'd like to add pps constraints to the throttling parameters (all it does is bandwidth right now). Hi Navdeep, After reviewing your patch, we've concluded that we can't use your flow-ID API's AS-IS for the mlxen hardware. We are working on a new patch proposal after the feedback received here, and will probably post something in August, hence now it is vacation time and not all people are available. Also we are worried that the m_tags cause un-needed overhead, that it invokes malloc for every pkt header duplication on transmit. As to give you some clues: One of FreeBSD's targets is firewalls and routers. We think that a flow ID / hardware queue feature should also be usable by the firewall, and not only limited to TCP/UDP. That means you can create queues for multiple connections which then get rate limited as a firewall rule. Right now our main target is not the firewall, but we see that by minor modifications of the APIs, this feature becomes very easy to implement. --HPS ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: [RFC] Add support for changing the flow ID of TCP connections
On Wed, Jul 09, 2014 at 04:36:53PM +0200, Hans Petter Selasky wrote: > On 07/08/14 21:17, Navdeep Parhar wrote: ... > > > >I think we need to design this to be as generic as possible. I have > >quite a bit of code that does this stuff but I haven't pushed it > >upstream or even offered it for review (yet). > > > > Hi, > > When will the non hardware related patches be available for review? > I understand there are multiple ways to reach the same goal, and I > think it would be great if we could agree on a common API for > applications. Here is the kernel side of the patch: http://people.freebsd.org/~np/flow_pacing_kern.diff The registration parameters and the throttling parameters are probably cxgbe-centric, because that's what it was written for. We'll need to tidy up those structs certainly. And I'd like to add pps constraints to the throttling parameters (all it does is bandwidth right now). Regards, Navdeep ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: [RFC] Add support for changing the flow ID of TCP connections
On 07/08/14 21:17, Navdeep Parhar wrote: On 07/08/14 10:46, Hans Petter Selasky wrote: Hi, I'm working on a new feature which will allow TCP connections to be timing controlled by the ethernet hardware driver, actually the mlxen driver. The main missing piece in the kernel is to allow the mbuf's flowid value to be overwritten in "struct inpcb" once the connection is established and to have a callback once the TCP connection is gone so that the assigned "flowid" can be freed by the ethernet hardware driver. The "flowid" will be used to assign the outgoing data traffic of a specific TCP connections to a hardware controlled queue, which in advance contain certain parameters about the timing for the transmitted packets. To be able to set the flowid I'm using existing functions in the kernel TCP code to lookup the "inpcb" structure based on the 4-tuple, via the "ifp->if_ioctl()" callback of the network adapter. I'm also registering a function method table so that I get a callback when the TCP connection is gone. A this point of development I would like to get some feedback from FreeBSD network guys about my attached patch proposal. The motivation for this work is to have a more reliable TCP transmissions typically for fixed-rate media content going some distance. To illustrate this I will give you an example from the world of VoIP, which is using UDP. When doing long-distance VoIP calls through various unknown networks and routers it makes a very big difference if you are sending data 20ms apart or 40ms apart, even at the exact same rate. In the one case you might experience a bunch of packet drops, and in the other case, everything is fine. Why? Because the number of packets you send per second, and the timing is important. The goal is to apply some timing rules for TCP, to increase the factor of successful transmission, and to reduce the amount of data loss. For high throughput applications we want to do this by means of hardware. While at it I would like to "typedef" the flowid used by mbufs, "struct inpcb" and many more places. Where would the right place be to put such a definition? In "sys/mbuf.h"? Comments are appreciated! I think we need to design this to be as generic as possible. I have quite a bit of code that does this stuff but I haven't pushed it upstream or even offered it for review (yet). Hi, When will the non hardware related patches be available for review? I understand there are multiple ways to reach the same goal, and I think it would be great if we could agree on a common API for applications. --HPS ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: [RFC] Add support for changing the flow ID of TCP connections
Hi! The flowid value has way, way too many possible meanings but it's always been a mostly-static value. I'm worried about overriding it with multiple meanings that cause features to not work at all together. So I'd rather leave the flowid/flowtype as it currently is so it doesn't upset packet reordering and can be used by things like RSS for scaling, and instead introduce a new connection ID to be used for your purpose. That way the existing use of flowid for packet ordering and flowid/flowtype for doing network scaling and netisr selection can work together with your connection id requirements. Having stack support for hardware/firmware packet scheduling is cool. It seems to somewhat overlap with other parts of the TCP offload though and I'm concerned about bloating out inpcb by 3 pointers for each connection where lots of connections on the same NIC will point to the same function set or NULL. I'd hit up what others in this space are doing. There's pacing support in the chelsio NIC for example and I'm not sure what Navdeep's plans are for that in upstream FreeBSD. Other than that, cool! -a On 8 July 2014 10:46, Hans Petter Selasky wrote: > Hi, > > I'm working on a new feature which will allow TCP connections to be timing > controlled by the ethernet hardware driver, actually the mlxen driver. The > main missing piece in the kernel is to allow the mbuf's flowid value to be > overwritten in "struct inpcb" once the connection is established and to have > a callback once the TCP connection is gone so that the assigned "flowid" can > be freed by the ethernet hardware driver. > > The "flowid" will be used to assign the outgoing data traffic of a specific > TCP connections to a hardware controlled queue, which in advance contain > certain parameters about the timing for the transmitted packets. > > To be able to set the flowid I'm using existing functions in the kernel TCP > code to lookup the "inpcb" structure based on the 4-tuple, via the > "ifp->if_ioctl()" callback of the network adapter. I'm also registering a > function method table so that I get a callback when the TCP connection is > gone. > > A this point of development I would like to get some feedback from FreeBSD > network guys about my attached patch proposal. > > The motivation for this work is to have a more reliable TCP transmissions > typically for fixed-rate media content going some distance. To illustrate > this I will give you an example from the world of VoIP, which is using UDP. > When doing long-distance VoIP calls through various unknown networks and > routers it makes a very big difference if you are sending data 20ms apart or > 40ms apart, even at the exact same rate. In the one case you might > experience a bunch of packet drops, and in the other case, everything is > fine. Why? Because the number of packets you send per second, and the timing > is important. The goal is to apply some timing rules for TCP, to increase > the factor of successful transmission, and to reduce the amount of data > loss. For high throughput applications we want to do this by means of > hardware. > > > While at it I would like to "typedef" the flowid used by mbufs, "struct > inpcb" and many more places. Where would the right place be to put such a > definition? In "sys/mbuf.h"? > > > Comments are appreciated! > > --HPS > > ___ > freebsd-current@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-current > To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org" ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: [RFC] Add support for changing the flow ID of TCP connections
On 07/08/14 10:46, Hans Petter Selasky wrote: > Hi, > > I'm working on a new feature which will allow TCP connections to be > timing controlled by the ethernet hardware driver, actually the mlxen > driver. The main missing piece in the kernel is to allow the mbuf's > flowid value to be overwritten in "struct inpcb" once the connection is > established and to have a callback once the TCP connection is gone so > that the assigned "flowid" can be freed by the ethernet hardware driver. > > The "flowid" will be used to assign the outgoing data traffic of a > specific TCP connections to a hardware controlled queue, which in > advance contain certain parameters about the timing for the transmitted > packets. > > To be able to set the flowid I'm using existing functions in the kernel > TCP code to lookup the "inpcb" structure based on the 4-tuple, via the > "ifp->if_ioctl()" callback of the network adapter. I'm also registering > a function method table so that I get a callback when the TCP connection > is gone. > > A this point of development I would like to get some feedback from > FreeBSD network guys about my attached patch proposal. > > The motivation for this work is to have a more reliable TCP > transmissions typically for fixed-rate media content going some > distance. To illustrate this I will give you an example from the world > of VoIP, which is using UDP. When doing long-distance VoIP calls through > various unknown networks and routers it makes a very big difference if > you are sending data 20ms apart or 40ms apart, even at the exact same > rate. In the one case you might experience a bunch of packet drops, and > in the other case, everything is fine. Why? Because the number of > packets you send per second, and the timing is important. The goal is to > apply some timing rules for TCP, to increase the factor of successful > transmission, and to reduce the amount of data loss. For high throughput > applications we want to do this by means of hardware. > > > While at it I would like to "typedef" the flowid used by mbufs, "struct > inpcb" and many more places. Where would the right place be to put such > a definition? In "sys/mbuf.h"? > > > Comments are appreciated! I think we need to design this to be as generic as possible. I have quite a bit of code that does this stuff but I haven't pushed it upstream or even offered it for review (yet). cxgbe(4) hardware does throttling and traffic pacing too, but it's not limited to TCP, and it can do it per queue or per "flow" -- you can limit a tx queue or an individual flow to a packet-per-second limit or a bandwidth ceiling; this works for both plain NIC (TCP, UDP, whatever), as well as stateful TCP offload). For TCP (NIC or TOE) the chip can even rewrite the TCP timestamp to account for the extra time that the chip/driver held the packet because it was asked to slow down a flow. The per queue stuff is handled via a driver-specific tool (cxgbetool). For per-flow throttling my implementation adds a new sockopt (SO_TX_THROTTLE) that lets an application specify a throttle rate for a socket. The kernel allocates a "flow identifier" for each such socket and tcp_output (or udp_output, ..) will attach an mbuf tag containing this identifier and throttling parameters to each mbuf that it pushes out. Drivers for hardware that can throttle traffic look for this tag, the rest ignore it. - cxgbe(4) registers itself as a "flow throttling provider" with the kernel when it attaches to the chip. It tells the kernel how many flows it can handle and the range of rates it can handle. - setsockopt(SO_TX_THROTTLE, rate) makes the kernel allocate a unique identifier for the socket. This is *not* related to the RSS flowid at all. If a listening socket has SO_TX_THROTTLE, all its children will inherit the rate limiting parameters but will each get its own unique identifier. The setsockopt fails if there aren't any flow throttling providers registered, - tcp_output (and other proto_output) routines look for SO_TX_THROTTLE and attach extra metadata, in the form of a tag, to the outgoing frames. - cxgbe(4) reads this metadata and acts on it. Regards, Navdeep ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"