Re: [ovs-dev] [PATCH 16/17] dpif-netdev: Centralized threads and queues handling code.

2016-11-23 Thread Bodireddy, Bhanuprakash
>-Original Message- >From: Daniele Di Proietto [mailto:diproiet...@ovn.org] >Sent: Tuesday, November 22, 2016 9:01 PM >To: Bodireddy, Bhanuprakash <bhanuprakash.bodire...@intel.com> >Cc: d...@openvswitch.org >Subject: Re: [ovs-dev] [PATCH 16/17] dpif-netdev: Central

Re: [ovs-dev] [PATCH 16/17] dpif-netdev: Centralized threads and queues handling code.

2016-11-22 Thread Bodireddy, Bhanuprakash
>-Original Message- >From: ovs-dev-boun...@openvswitch.org [mailto:ovs-dev- >boun...@openvswitch.org] On Behalf Of Daniele Di Proietto >Sent: Wednesday, November 16, 2016 12:46 AM >To: d...@openvswitch.org >Cc: Daniele Di Proietto >Subject: [ovs-dev] [PATCH 16/17]

Re: [ovs-dev] [PATCH v3 00/12] Improve performance of OVS-DPDK classifier.

2016-11-15 Thread Bodireddy, Bhanuprakash
>-Original Message- >From: Daniele Di Proietto [mailto:diproiet...@ovn.org] >Sent: Monday, November 14, 2016 10:45 PM >To: Bodireddy, Bhanuprakash <bhanuprakash.bodire...@intel.com> >Cc: d...@openvswitch.org; Jarno Rajahalme <ja...@ovn.org> >Subject: Re: [ovs

Re: [ovs-dev] [PATCH] netdev-dpdk: Use intermediate queue during packet transmission.

2016-12-20 Thread Bodireddy, Bhanuprakash
>-Original Message- >From: Ilya Maximets [mailto:i.maxim...@samsung.com] >Sent: Tuesday, December 20, 2016 8:09 AM >To: Bodireddy, Bhanuprakash <bhanuprakash.bodire...@intel.com>; Aaron >Conole <acon...@redhat.com> >Cc: d...@openvswitch.org; Daniele Di Pr

Re: [ovs-dev] [PATCH] netdev-dpdk: Use intermediate queue during packet transmission.

2016-12-19 Thread Bodireddy, Bhanuprakash
Thanks Ilya and Aaron for reviewing this patch and providing your comments, my reply inline. >-Original Message- >From: Ilya Maximets [mailto:i.maxim...@samsung.com] >Sent: Monday, December 19, 2016 8:41 AM >To: Aaron Conole <acon...@redhat.com>; Bodir

Re: [ovs-dev] Ports down on OVS DPDK and traffic fails

2017-03-30 Thread Bodireddy, Bhanuprakash
>-Original Message- >From: ovs-dev-boun...@openvswitch.org [mailto:ovs-dev- >boun...@openvswitch.org] On Behalf Of Nadathur, Sundar >Sent: Thursday, March 30, 2017 6:46 PM >To: ovs-dev@openvswitch.org >Subject: [ovs-dev] Ports down on OVS DPDK and traffic fails > >Summary: I am trying to

Re: [ovs-dev] [PATCH 3/7] netdev-dpdk: Add support for keepalive functionality.

2017-04-03 Thread Bodireddy, Bhanuprakash
>>> >>>This whole mechanism seems very error prone. Is it possible to hang a >>>thread with the subsequent sem_post? >> >> The relay function is called by 'ovs_keepalive' thread. I didn't >> completely understand your concerns here. I would be happy to verify >> if you have any scenarios in mind

Re: [ovs-dev] [PATCH 7/7] Documentation: Update DPDK doc with Keepalive feature.

2017-04-03 Thread Bodireddy, Bhanuprakash
>-Original Message- >From: ovs-dev-boun...@openvswitch.org [mailto:ovs-dev- >boun...@openvswitch.org] On Behalf Of Stephen Finucane >Sent: Monday, April 3, 2017 11:15 AM >To: ovs-dev@openvswitch.org >Subject: Re: [ovs-dev] [PATCH 7/7] Documentation: Update DPDK doc with >Keepalive feature.

Re: [ovs-dev] [PATCH 5/7] utils: Introduce xusleep for subsecond granularity.

2017-04-03 Thread Bodireddy, Bhanuprakash
>Bhanuprakash Bodireddy writes: > >> This will be used by KA framework that needs millisecond granularity. >> >> Signed-off-by: Bhanuprakash Bodireddy >> >> --- > >Without this patch, builds starting at 3/7 will fail. That's

Re: [ovs-dev] [PATCH 0/7] Add OVS DPDK keep-alive functionality

2017-04-03 Thread Bodireddy, Bhanuprakash
Thanks Aaron for reviewing this patch series. My comments inline. > >Hi Bhanu, > >Bhanuprakash Bodireddy writes: > >> This patch is aimed at achieving Fastpath Service Assurance in >> OVS-DPDK deployments. This commit adds support for monitoring the >> packet

Re: [ovs-dev] [PATCH 1/7] dpdk: Add helper functions for DPDK keepalive.

2017-04-03 Thread Bodireddy, Bhanuprakash
>-Original Message- >From: Aaron Conole [mailto:acon...@redhat.com] >Sent: Monday, April 3, 2017 1:58 AM >To: Bodireddy, Bhanuprakash <bhanuprakash.bodire...@intel.com> >Cc: d...@openvswitch.org >Subject: Re: [ovs-dev] [PATCH 1/7] dpdk: Add helper funct

Re: [ovs-dev] [PATCH 3/7] netdev-dpdk: Add support for keepalive functionality.

2017-04-03 Thread Bodireddy, Bhanuprakash
>> >> +/* >> + * OVS Shared Memory structure >> + * >> + * The information in the shared memory block will be read by collectd. >> + * */ >> +struct dpdk_keepalive_shm { >> +/* IPC semaphore. Posted when a core dies */ >> +sem_t core_died; >> + >> +/* >> + * Relayed status of each

Re: [ovs-dev] [PATCH 4/7] process: Retrieve process status.

2017-04-03 Thread Bodireddy, Bhanuprakash
>-Original Message- >From: Aaron Conole [mailto:acon...@redhat.com] >Sent: Monday, April 3, 2017 2:00 AM >To: Bodireddy, Bhanuprakash <bhanuprakash.bodire...@intel.com> >Cc: d...@openvswitch.org >Subject: Re: [ovs-dev] [PATCH 4/7] process: Retrieve process status. &

Re: [ovs-dev] [PATCH 1/5] dpif-netdev: Skip EMC lookup when EMC is disabled.

2017-04-13 Thread Bodireddy, Bhanuprakash
>On 04/13/2017 07:11 PM, Kevin Traynor wrote: >> On 03/12/2017 05:33 PM, Bhanuprakash Bodireddy wrote: >>> Conditional EMC insert patch gives the flexibility to configure the >>> probability of flow insertion in to EMC. This also allows an option >>> to entirely disable EMC by setting

Re: [ovs-dev] [PATCH v3] netdev-dpdk: Implement Tx intermediate queue for dpdk ports.

2017-04-14 Thread Bodireddy, Bhanuprakash
(ns) Regards, Bhanuprakash. >-Original Message- >From: Bodireddy, Bhanuprakash >Sent: Thursday, February 2, 2017 10:15 PM >To: d...@openvswitch.org >Cc: i.maxim...@samsung.com; ktray...@redhat.com; diproiet...@ovn.org; >Bodireddy, Bhanuprakash <bhanuprakash.bodire...@

Re: [ovs-dev] [PATCH 4/5] dpif-netdev: Fix comments for dp_netdev_pmd_thread struct.

2017-04-18 Thread Bodireddy, Bhanuprakash
>On Sun, Mar 12, 2017 at 05:33:27PM +, Bhanuprakash Bodireddy wrote: >> The sorted subtable ranking patch introduced a classifier instance per >> ingress port with its subtables ranked on the frequency of hits. The >> pmd thread can have more classifier instances now and solely depends >> on

Re: [ovs-dev] [PATCH v3] netdev-dpdk: Implement Tx intermediate queue for dpdk ports.

2017-04-18 Thread Bodireddy, Bhanuprakash
Hi Eelco, Please find my comments inline. > >Hi Bhanuprakash, > >I was doing some Physical to Virtual tests, and whenever the number of flows >reaches the rx batch size performance dropped a lot. I created an >experimental patch where I added an intermediate queue and flush it at the >end of

Re: [ovs-dev] [PATCH 3/7] netdev-dpdk: Add support for keepalive functionality.

2017-04-24 Thread Bodireddy, Bhanuprakash
>>> The relay function is called by 'ovs_keepalive' thread. I didn't >>> completely understand your concerns here. I would be happy to verify >>> if you have any scenarios in mind that you think pose problems. >> >>My concern stems from this: >> "is relayed to monitoring framework by unlocking

Re: [ovs-dev] [PATCH 0/4] prioritizing latency sensitive traffic

2017-07-28 Thread Bodireddy, Bhanuprakash
Hi Billy, >Hi All, > >This patch set provides a method to request ingress scheduling on interfaces. >It also provides an implemtation of same for DPDK physical ports. > >This allows specific packet types to be: >* forwarded to their destination port ahead of other packets. >and/or >* be less

Re: [ovs-dev] [PATCH v4 5/5] dpif-netdev: Flush the packets in intermediate queue.

2017-08-11 Thread Bodireddy, Bhanuprakash
. Bhanuprakash. >-Original Message- >From: Darrell Ball [mailto:db...@vmware.com] >Sent: Friday, August 11, 2017 2:03 AM >To: Bodireddy, Bhanuprakash <bhanuprakash.bodire...@intel.com>; >d...@openvswitch.org >Subject: Re: [ovs-dev] [PATCH v4 5/5] dpif-n

Re: [ovs-dev] [PATCH v4 2/5] netdev-dpdk: Add netdev_dpdk_vhost_txq_flush function.

2017-08-11 Thread Bodireddy, Bhanuprakash
>On 09.08.2017 15:35, Bodireddy, Bhanuprakash wrote: >>>> >>>> +static int >>>> +netdev_dpdk_vhost_tx_burst(struct netdev_dpdk *dev, int qid) { >>>> +struct dpdk_tx_queue *txq = >tx_q[qid]; >>>> +struct rte

Re: [ovs-dev] [PATCH v4 2/5] netdev-dpdk: Add netdev_dpdk_vhost_txq_flush function.

2017-08-10 Thread Bodireddy, Bhanuprakash
>> } else { +/* If the queue is disabled in the guest, the corresponding qid + * map shall be set to OVS_VHOST_QUEUE_DISABLED(-2). + * + * The packets that were queued in 'qid' could be

Re: [ovs-dev] [PATCH v2 1/2] util: Make PADDED_MEMBERS work more than once per struct.

2017-07-13 Thread Bodireddy, Bhanuprakash
>Until now, if the PADDED_MEMBERS macro was used more than once in a >struct, it caused Clang and GCC warnings because of multiple definition of a >member named "pad". This commit fixes the problem by giving each of these >a unique name. > >MSVC, Clang, and GCC all support the __COUNTER__ macro,

Re: [ovs-dev] [PATCH] packets: Do not initialize ct_orig_tuple.

2017-07-12 Thread Bodireddy, Bhanuprakash
Thanks Darrell for testing this patch. I will send on v2 with your suggestions below. - Bhanuprakash. >-Original Message- >From: Darrell Ball [mailto:db...@vmware.com] >Sent: Sunday, July 9, 2017 9:44 PM >To: Bodireddy, Bhanuprakash <bhanuprakash.bodire...

Re: [ovs-dev] [PATCH] packets: Reorganize the pkt_metdata structure.

2017-07-12 Thread Bodireddy, Bhanuprakash
>On Sun, Jul 09, 2017 at 10:53:52PM +, Darrell Ball wrote: >> I went thru. this patch and see the merits of the objective. >> I also did various testing with it. >> I had one comment inline. Thanks Darrell for testing this patch. I will factor in your comment in next version of patch. >>

Re: [ovs-dev] [PATCH] packets: Reorganize the pkt_metdata structure.

2017-07-12 Thread Bodireddy, Bhanuprakash
Hi Ben, >On Thu, Jun 22, 2017 at 10:10:49PM +0100, Bhanuprakash Bodireddy wrote: >> pkt_metadata_init() is called for every packet in userspace datapath >> and initializes few members in pkt_metadata. Before this the members >> that needs to be initialized are prefetched using

Re: [ovs-dev] [PATCH] packets: Reorganize the pkt_metdata structure.

2017-07-12 Thread Bodireddy, Bhanuprakash
> >On Wed, Jul 12, 2017 at 10:38:30AM +, Bodireddy, Bhanuprakash wrote: >> Hi Ben, >> >> >On Thu, Jun 22, 2017 at 10:10:49PM +0100, Bhanuprakash Bodireddy wrote: >> >> pkt_metadata_init() is called for every packet in userspace >> >> datapath

Re: [ovs-dev] [PATCH 00/20] Add OVS DPDK keep-alive functionality.

2017-07-21 Thread Bodireddy, Bhanuprakash
>> Keepalive feature is aimed at achieving Fastpath Service Assurance in >> OVS-DPDK deployments. It adds support for monitoring the packet >> processing cores(PMD thread cores) by dispatching heartbeats at >> regular intervals. Incase of heartbeat misses additional health checks >> are enabled on

Re: [ovs-dev] [PATCH] packets: Reorganize the pkt_metdata structure.

2017-07-24 Thread Bodireddy, Bhanuprakash
Hello Ben, >> > >> >I doubt that this is about a warning, because as I understand it OVS >> >on MSVC causes a lot of warnings, so it's probably a more serious issue. >> >> In will try to get MSVC installed and verify the OvS build. Will get back to >> you >on this. > >If you push to a branch on

Re: [ovs-dev] [PATCH v2 0/4] Output packet batching.

2017-07-27 Thread Bodireddy, Bhanuprakash
HI Ilya, I am OOO and would review and test this patch series shortly(by Monday). Bhanuprakash. >-Original Message- >From: Ilya Maximets [mailto:i.maxim...@samsung.com] >Sent: Wednesday, July 26, 2017 4:21 PM >To: ovs-dev@openvswitch.org; Bodireddy, Bhanuprakash ><b

Re: [ovs-dev] [PATCH 5/6] netdev-dpdk: Add netdev_dpdk_vhost_txq_flush function.

2017-06-28 Thread Bodireddy, Bhanuprakash
> >On 27.06.2017 23:31, Bodireddy, Bhanuprakash wrote: >>> On 26.06.2017 00:52, Bodireddy, Bhanuprakash wrote: >>>>>> + >>>>>> +/* Flush the txq if there are any packets available. >>>>>

Re: [ovs-dev] [ovs-dev, 3/6] netdev-dpdk: Add intermediate queue support.

2017-06-28 Thread Bodireddy, Bhanuprakash
>At first, this patch should be applied after the patch with flushing on >reconfiguration because we must not reconfigure ports while there are >unsent packets in the intermediate queue. >Otherwise we may destroy the memory pool which contains that packets and >will try to send them after that.

Re: [ovs-dev] [PATCH 0/3] Output packet batching.

2017-07-04 Thread Bodireddy, Bhanuprakash
Apologies for snipping the text. I did it to keep this thread readable. > >Hi Darrell and Jan. >Thanks for looking at this. I agree with Darrell that mixing implementations on >two different levels is a bad idea, but as I already wrote in reply to >Bhanuprakash [2], there is no issues with

Re: [ovs-dev] [PATCH 0/3] Output packet batching.

2017-07-03 Thread Bodireddy, Bhanuprakash
>This patch-set inspired by [1] from Bhanuprakash Bodireddy. >Implementation of [1] looks very complex and introduces many pitfalls for >later code modifications like possible packet stucks. > >This version targeted to make simple and flexible output packet batching on >higher level without

Re: [ovs-dev] [ovs-dev, 4/6] dpif-netdev: Flush the packets in intermediate queue.

2017-06-29 Thread Bodireddy, Bhanuprakash
>I found another issue while testing w.r.t vhostuser ports. > >When non-pmd thread tries to send packets to vHostUser port below is the >call patch > dp_execute_cb() > netdev_dpdk_vhost_send() [(may_steal == true) and (Pkt src type == >DPBUF_MALLOC)] >dpdk_do_tx_copy() >

Re: [ovs-dev] [ovs-dev, 4/6] dpif-netdev: Flush the packets in intermediate queue.

2017-06-28 Thread Bodireddy, Bhanuprakash
>At least, you have to flush non-PMD threads too. In case of non PMD threads we don’t have to flush as the packets aren't queued and bursted instantly. The call path on the transmit side is: Vswitchd thread: dp_execute_cb() netdev_send()

Re: [ovs-dev] [PATCH v2] process: Consolidate process related APIs.

2017-06-29 Thread Bodireddy, Bhanuprakash
. >-Original Message- >From: Bodireddy, Bhanuprakash >Sent: Tuesday, June 20, 2017 10:30 AM >To: d...@openvswitch.org >Cc: Bodireddy, Bhanuprakash <bhanuprakash.bodire...@intel.com>; Ben >Pfaff <b...@ovn.org> >Subject: [PATCH v2] process: Consolidate process relat

Re: [ovs-dev] [PATCH 5/6] netdev-dpdk: Add netdev_dpdk_vhost_txq_flush function.

2017-06-27 Thread Bodireddy, Bhanuprakash
>On 26.06.2017 00:52, Bodireddy, Bhanuprakash wrote: >>>> + >>>> +/* Flush the txq if there are any packets available. >>>> + * dynamic_txqs/concurrent_txq is disabled for vHost User ports as >>>> + * 'OVS_VHOST_MAX_QUEUE_NUM' txqs are preallo

Re: [ovs-dev] [PATCH v2 6/6] Documentation: Update DPDK doc with Keepalive feature.

2017-04-27 Thread Bodireddy, Bhanuprakash
>> >>    Compute NodeControllerCompute Node >> >> Collectd  <--> Ceilometer <>   Collectd >> >> OvS DPDK   OvS DPDK >> >>    +-+ >>    | VM  | >>    +--+--+ >>   \---+---/ >>   | >>    +--+---+   

Re: [ovs-dev] [PATCH 4/5] dpif-netdev: Fix comments for dp_netdev_pmd_thread struct.

2017-04-28 Thread Bodireddy, Bhanuprakash
> > >How about something like this ? > >+ * Each struct has its own flow cache and classifier per managed inport. >Packets >+ *received from managed ports are looked up in the corresponding pmd >+ * thread's flow cache and corresponding classifier, if the flow cache >misses. >+ *

Re: [ovs-dev] [PATCH 0/7] Add OVS DPDK keep-alive functionality

2017-04-24 Thread Bodireddy, Bhanuprakash
>> > Agree that Latency is indeed important. I would collect the latency >> > stats and will share the results in this thread. >> >> Awesome. Even better would be to put those informations in the cover >> letter of v2. Just a short summary of the tests and what the results >> were for latency /

Re: [ovs-dev] [PATCH v2 0/6] Add OVS DPDK keep-alive functionality

2017-04-28 Thread Bodireddy, Bhanuprakash
>> This patch is aimed at achieving Fastpath Service Assurance in >> OVS-DPDK deployments. This commit adds support for monitoring the >> packet processing cores(pmd thread cores) by dispatching heartbeats at >> regular intervals. Incase of heartbeat miss the failure shall be >> detected &

Re: [ovs-dev] [PATCH v3 0/6] netdev-dpdk: Use intermediate queue during packet transmission.

2017-08-08 Thread Bodireddy, Bhanuprakash
Hi Darrell, >Sorry, I was multitasking last week and did not get a chance to finish the >responses on Friday > >I looked thru. the code for all the patches The last 3 patches of V3 needed a >manual merge; as you know, the series needs a rebase after recent commits. I will rebase and send out

Re: [ovs-dev] [PATCH v3 2/6] netdev-dpdk: Add netdev_dpdk_txq_flush function.

2017-08-08 Thread Bodireddy, Bhanuprakash
>Hi Bhanu > >Would it be possible to combine patches 1 and 2, rather than initially defining >an empty netdev_txq_flush for dpdk ? I think the combined patch would have >more context. No problem Darrell . I will merge 1 & 2 in V4. - Bhanuprakash. > > >-Original Message- >From:

Re: [ovs-dev] [PATCH v3 4/6] netdev-dpdk: Add intermediate queue support.

2017-08-08 Thread Bodireddy, Bhanuprakash
> >This commit introduces netdev_dpdk_eth_tx_queue() function that >implements intermediate queue and packet buffering. The packets get >buffered till the threshold 'INTERIM_QUEUE_BURST_THRESHOLD[32] is >reached and eventually gets transmitted. > >To handle the case(eg: ping)

Re: [ovs-dev] [PATCH v4 1/5] netdev: Add netdev_txq_flush function.

2017-08-09 Thread Bodireddy, Bhanuprakash
Hi Ilya, >> >> +/* Flush tx queues. >> + * This is done periodically to empty the intermediate queue in case >> +of >> + * fewer packets (< INTERIM_QUEUE_BURST_THRESHOLD) buffered in the >queue. >> + */ >> +static int >> +netdev_dpdk_txq_flush(struct netdev *netdev, int qid , bool >>

Re: [ovs-dev] [PATCH v4 2/5] netdev-dpdk: Add netdev_dpdk_vhost_txq_flush function.

2017-08-09 Thread Bodireddy, Bhanuprakash
>> >> +static int >> +netdev_dpdk_vhost_tx_burst(struct netdev_dpdk *dev, int qid) { >> +struct dpdk_tx_queue *txq = >tx_q[qid]; >> +struct rte_mbuf **cur_pkts = (struct rte_mbuf >> +**)txq->vhost_burst_pkts; >> + >> +int tx_vid = netdev_dpdk_get_vid(dev); >> +int tx_qid = qid *

Re: [ovs-dev] [PATCH v4 2/5] netdev-dpdk: Add netdev_dpdk_vhost_txq_flush function.

2017-08-09 Thread Bodireddy, Bhanuprakash
>enable) if (enable) { dev->tx_q[qid].map = qid; >> >> Here flushing required too because we're possibly enabling previously >remapped queue. >> } else { +/* If the queue is disabled in the guest, the

Re: [ovs-dev] [PATCH v2 0/4] Output packet batching.

2017-08-01 Thread Bodireddy, Bhanuprakash
>This patch-set inspired by [1] from Bhanuprakash Bodireddy. >Implementation of [1] looks very complex and introduces many pitfalls for >later code modifications like possible packet stucks. > >This version targeted to make simple and flexible output packet batching on >higher level without

Re: [ovs-dev] [PATCH RFC v2 4/4] dpif-netdev: Time based output batching.

2017-08-01 Thread Bodireddy, Bhanuprakash
>On 28.07.2017 10:20, Darrell Ball wrote: >> I have not tested yet >> >> However, I would have expected something max latency config. to be >specific to netdev-dpdk port types > >IMHO, if we can make it generic, we must make it generic. > >[Darrell] >The first question I ask

Re: [ovs-dev] DPDK Merge Repo

2017-08-02 Thread Bodireddy, Bhanuprakash
>> Hi Darrell and Ben. >> >> > Hi All >> > >> > As mentioned before, I am using a repo for DPDK patch merging. >> > The repo is here: >> > https://github.com/darball/ovs/ >> > >> > There are still some outstanding patches from Bhanu that have not >> completed review yet: >> > >> > util: Add

Re: [ovs-dev] [PATCH v3 6/6] dpif-netdev: Flush the packets in intermediate queue.

2017-08-08 Thread Bodireddy, Bhanuprakash
Hi Darrell, > >Under low rate traffic conditions, there can be 2 issues. > (1) Packets potentially can get stuck in the intermediate queue. > (2) Latency of the packets can increase significantly due to > buffering in intermediate queue. > >This commit handles the (1)

Re: [ovs-dev] [PATCH v3 00/19] Add OVS DPDK keep-alive functionality.

2017-08-08 Thread Bodireddy, Bhanuprakash
by removing the dependency on DPDK Keepalive library it might be even better. I will work on this and wait for inputs from other reviewers before posting next version. - Bhanuprakash. > >Best regards, Ilya Maximets. > >On 04.08.2017 18:24, Bodireddy, Bhanuprakash wrote:

Re: [ovs-dev] [PATCH v3 00/19] Add OVS DPDK keep-alive functionality.

2017-08-04 Thread Bodireddy, Bhanuprakash
e future. Regards, Bhanuprakash. >-Original Message- >From: Ilya Maximets [mailto:i.maxim...@samsung.com] >Sent: Friday, August 4, 2017 2:40 PM >To: ovs-dev@openvswitch.org; Bodireddy, Bhanuprakash ><bhanuprakash.bodire...@intel.com> >Cc: Darrell Ball <db...@v

Re: [ovs-dev] [PATCH v2 0/6] Add OVS DPDK keep-alive functionality

2017-05-01 Thread Bodireddy, Bhanuprakash
>> > This patch is aimed at achieving Fastpath Service Assurance in >> > OVS-DPDK deployments. This commit adds support for monitoring the >> > packet processing cores(pmd thread cores) by dispatching heartbeats >> > at regular intervals. Incase of heartbeat miss the failure shall be >> > detected

Re: [ovs-dev] [PATCH v2 0/6] Add OVS DPDK keep-alive functionality

2017-05-01 Thread Bodireddy, Bhanuprakash
>There two ways to go with the design. > >1) Make it generic, so that it is not so PMD specific, as it is now. >2) If it stays PMD specific, make it stronger; right now, the health check is >limited – it detects that a PMD thread is proceeding or not. >For something like DPDK, I don’t think that

Re: [ovs-dev] [RFC] packets: Do not initialize ct_orig_tuple.

2017-05-15 Thread Bodireddy, Bhanuprakash
>Commit daf4d3c18da4("odp: Support conntrack orig tuple key.") introduced >new fields in struct 'pkt_metadata'. pkt_metadata_init() is called for every >packet in the userspace datapath. When testing a simple single flow case with >DPDK, we observe a lower throughput after the above commit (it

Re: [ovs-dev] [RFC PATCH] netdev-dpdk: Add Tx intermediate queue for vhost ports.

2017-05-15 Thread Bodireddy, Bhanuprakash
Hi Eelco, > > This commit adds the intermediate queue for vHost-user ports. It > >improves the throughput in multiple virtual machines deployments and > also >in cases with VM doing packet forwarding in kernel stack. > > > > This patch is aligned with intermediate queue implementation for dpdk

Re: [ovs-dev] [PATCH 4/5] dpif-netdev: Fix comments for dp_netdev_pmd_thread struct.

2017-05-15 Thread Bodireddy, Bhanuprakash
>> >> >>How about something like this ? >> >>+ * Each struct has its own flow cache and classifier per managed inport. >>Packets >>+ *received from managed ports are looked up in the corresponding pmd >>+ * thread's flow cache and corresponding classifier, if the flow >>cache misses.

Re: [ovs-dev] [PATCH 1/6] process: Consolidate process related APIs.

2017-06-20 Thread Bodireddy, Bhanuprakash
Hi Ben, >On Mon, Jun 19, 2017 at 07:53:59PM +0100, Bhanuprakash Bodireddy wrote: >> As part of retrieving system statistics, process status APIs along >> with helper functions were implemented. Some of them are very generic >> and can be reused by other subsystems. >> >> Move the APIs in

Re: [ovs-dev] [PATCH v10] netdev-dpdk: Increase pmd thread priority.

2017-06-26 Thread Bodireddy, Bhanuprakash
>With this change and CFS in effect, it effectively means that the dpdk control >threads need to be on different cores than the PMD threads or the response >latency may be too long for their control work ? >Have we tested having the control threads on the same cpu with -20 nice for >the pmd thread

Re: [ovs-dev] [PATCH v9] netdev-dpdk: Increase pmd thread priority.

2017-06-25 Thread Bodireddy, Bhanuprakash
>-Original Message- >From: nickcooper-zhangtonghao [mailto:n...@opencloud.tech] >Sent: Friday, June 23, 2017 4:00 PM >To: Bodireddy, Bhanuprakash <bhanuprakash.bodire...@intel.com> >Cc: d...@openvswitch.org >Subject: Re: [ovs-dev] [PATCH v9] netdev-dpdk: Incr

Re: [ovs-dev] [PATCH 5/6] netdev-dpdk: Add netdev_dpdk_vhost_txq_flush function.

2017-06-25 Thread Bodireddy, Bhanuprakash
>> + >> +/* Flush the txq if there are any packets available. >> + * dynamic_txqs/concurrent_txq is disabled for vHost User ports as >> + * 'OVS_VHOST_MAX_QUEUE_NUM' txqs are preallocated. >> + */ > >This comment is completely untrue. You may ignore 'concurrent_txq' >because you *must* lock the

Re: [ovs-dev] [PATCH 6/6] netdev: Fix null pointer dereference reported by clang.

2017-06-25 Thread Bodireddy, Bhanuprakash
Hi Mark, >> >>Clang reports that array access from 'dumps' variable result in null >>pointer dereference. >> >>Signed-off-by: Bhanuprakash Bodireddy >> > >Hi Bhanu, > >LGTM - I also compiled this with gcc, clang, and sparse without issue. >Checkpatch reports no

Re: [ovs-dev] [PATCH 7/8] netdev-dpdk: Configurable retries while enqueuing to vHost User ports.

2017-06-20 Thread Bodireddy, Bhanuprakash
>>On 06/07/2017 10:21 AM, Bhanuprakash Bodireddy wrote: >>> This commit adds "vhost-enque-retry" where in the number of retries >>> performed while enqueuing packets to vHostUser ports can be >>> configured in ovsdb. >>> >>> Currently number of retries are set to '8' and a retry is performed >>>

Re: [ovs-dev] [PATCH] doc: Resolve pep8 warnings in conf.py

2017-05-19 Thread Bodireddy, Bhanuprakash
>-Original Message- >From: ovs-dev-boun...@openvswitch.org [mailto:ovs-dev- >boun...@openvswitch.org] On Behalf Of Stephen Finucane >Sent: Friday, May 19, 2017 10:15 AM >To: d...@openvswitch.org >Subject: [ovs-dev] [PATCH] doc: Resolve pep8 warnings in conf.py > >flake8 doesn't like us

Re: [ovs-dev] [RFC 1/2] doc: Reduce duplication in 'man_pages'

2017-05-19 Thread Bodireddy, Bhanuprakash
The make fails with below error on my fedora 22 Target. error.log-- Documentation/conf.py:126:9: F812 list comprehension redefines 'filename' from line 59 Makefile:5848: recipe for target 'flake8-check' failed make[2]: ***

Re: [ovs-dev] [PATCH v8] netdev-dpdk: Increase pmd thread priority

2017-05-26 Thread Bodireddy, Bhanuprakash
ase and send the patch. - Bhanuprakash. > >> -Original Message- >> From: ovs-dev-boun...@openvswitch.org [mailto:ovs-dev- >> boun...@openvswitch.org] On Behalf Of Aaron Conole >> Sent: Tuesday, January 3, 2017 8:08 PM >> To: Bodireddy, Bhanuprakash <bhanuprakash.bodire...

Re: [ovs-dev] [RFC PATCH 0/6] Change dpdk rxq scheduling to incorporate rxq processing cycles.

2017-05-22 Thread Bodireddy, Bhanuprakash
>> Thanks for working on this. This work is important and would also solve OvS- >DPDK scaling issue. >> I have reviewed the patch series but haven't tested this yet with traffic. I >would stick to comments on high level design at his stage. >> >> With this series, rxqs are sorted based on cycles

Re: [ovs-dev] [RFC PATCH 0/6] Change dpdk rxq scheduling to incorporate rxq processing cycles.

2017-05-21 Thread Bodireddy, Bhanuprakash
Hello Kevin, > >Rxqs are scheduled to be handled across available pmds in round robin order >with no weight or priority. > >It can happen that some very busy queues are handled by one pmd which >does not have enough cycles to prevent packets being dropped on them. >While at the same time another

Re: [ovs-dev] [RFC PATCH] netdev-dpdk: Add Tx intermediate queue for vhost ports.

2017-06-07 Thread Bodireddy, Bhanuprakash
Hi Eelco, >On 23/04/17 17:41, Bhanuprakash Bodireddy wrote: > > > This commit adds the intermediate queue for vHost-user ports. It > >improves the throughput in multiple virtual machines deployments and > also >in cases with VM doing packet forwarding in kernel stack. > > > > This patch is

Re: [ovs-dev] [PATCH v2 0/6] Add OVS DPDK keep-alive functionality

2017-06-07 Thread Bodireddy, Bhanuprakash
Hi Darrel, Aaron >There two ways to go with the design. > >1) Make it generic, so that it is not so PMD specific, as it is now. >2) If it stays PMD specific, make it stronger; right now, the health check is >limited – it detects that a PMD thread is proceeding or not. >For something like DPDK, I

Re: [ovs-dev] [RFC PATCH 02/21] ovsschema: Introduce 'keepalive' column in Open_vSwitch.

2017-06-08 Thread Bodireddy, Bhanuprakash
>On Wed, Jun 07, 2017 at 05:14:58PM +0100, Bhanuprakash Bodireddy wrote: >> This commit adds new ovsdb column "keepalive". It shows the overall >> datapath status and the health of the cores running datapath threads. >> >> Signed-off-by: Bhanuprakash Bodireddy >>

Re: [ovs-dev] [RFC PATCH v2 03/19] Keepalive: Add initial keepalive support.

2017-06-13 Thread Bodireddy, Bhanuprakash
>Hi Bhanu, > >Bhanuprakash Bodireddy writes: > >> This commit introduces the initial keepalive support by adding >> 'keepalive' module and also helper and initialization functions that >> will be invoked by later commits. >> >> This commit adds new ovsdb column

Re: [ovs-dev] [RFC PATCH v2 04/19] bridge: Invoke keepalive framework.

2017-06-13 Thread Bodireddy, Bhanuprakash
>Hi Bhanu, > >Bhanuprakash Bodireddy writes: > >> With this commit the keepalive framework is invoked when >> enable-keepalive is set to 'true' in ovsdb. As part of initialization >> the shared memory block is created and initialized. Also the keepalive >> data

Re: [ovs-dev] [PATCH 7/8] netdev-dpdk: Configurable retries while enqueuing to vHost User ports.

2017-06-13 Thread Bodireddy, Bhanuprakash
Hi Kevin, >On 06/07/2017 10:21 AM, Bhanuprakash Bodireddy wrote: >> This commit adds "vhost-enque-retry" where in the number of retries >> performed while enqueuing packets to vHostUser ports can be configured >> in ovsdb. >> >> Currently number of retries are set to '8' and a retry is performed

Re: [ovs-dev] [RFC PATCH 00/21] Add OVS DPDK keep-alive functionality

2017-06-14 Thread Bodireddy, Bhanuprakash
Hi Aaron, >Hi Bhanu, > >I've been playing with this a little bit; is it too late to consider tracking >'threads' >instead of 'cores'? I'm not sure what it means for a particular core ID to be >'healthy' - but I know what 'pmd24' not responding means. That's an interesting input. It's not late

Re: [ovs-dev] [PATCH 0/6 V2] netdev-dpdk: Use intermediate queue during packet transmission.

2017-06-19 Thread Bodireddy, Bhanuprakash
>-Original Message- >From: ovs-dev-boun...@openvswitch.org [mailto:ovs-dev- >boun...@openvswitch.org] On Behalf Of Bhanuprakash Bodireddy >Sent: Sunday, June 18, 2017 8:56 PM >To: d...@openvswitch.org >Subject: [ovs-dev] [PATCH 0/6 V2] netdev-dpdk: Use intermediate queue >during packet

Re: [ovs-dev] [RFC PATCH 00/21] Add OVS DPDK keep-alive functionality

2017-06-19 Thread Bodireddy, Bhanuprakash
Hi Aaron, >> >>I've been playing with this a little bit; is it too late to consider tracking >'threads' >>instead of 'cores'? I'm not sure what it means for a particular core >>ID to be 'healthy' - but I know what 'pmd24' not responding means. > >That's an interesting input. It's not late and

Re: [ovs-dev] [PATCH 0/8] netdev-dpdk: Use intermediate queue during packet transmission.

2017-06-13 Thread Bodireddy, Bhanuprakash
Hi Eelco >Hi Bhanu, > >Went over the full patch set, and the changes look good to me. >All my previous concerns are addressed, and therefore I'm acking this series. Thanks for reviewing the series and acking it. > >I do have one small remark regarding the dpdk_tx_queue struct, see >individual

Re: [ovs-dev] [PATCH v4 3/7] dpif-netdev: Register packet processing cores to KA framework.

2017-09-13 Thread Bodireddy, Bhanuprakash
>"Bodireddy, Bhanuprakash" <bhanuprakash.bodire...@intel.com> writes: > >>>Bhanuprakash Bodireddy <bhanuprakash.bodire...@intel.com> writes: >>> >>>> This commit registers the packet processing PMD cores to keepalive >>

[ovs-dev] ovs-tcpdump error

2017-09-21 Thread Bodireddy, Bhanuprakash
Hi, ovs-tcpdump throws the below error when trying to capture packets on one of the vhostuserports. $ ovs-tcpdump -i dpdkvhostuser0 ERROR: Please create an interface called `midpdkvhostuser0` See your OS guide for how to do this. Ex: ip link add midpdkvhostuser0 type veth peer name

Re: [ovs-dev] [PATCH 02/13] netdev-dummy: Reorder elements in dummy_packet_stream structure.

2017-09-18 Thread Bodireddy, Bhanuprakash
Hi greg, >On 09/08/2017 10:59 AM, Bhanuprakash Bodireddy wrote: >> By reordering elements in dummy_packet_stream structure, sum holes > >Do you mean "the sum of the holes" can be reduced or do you mean "some >holes" >can be reduced? In this patch series "sum of the holes" means, the sum/total of

Re: [ovs-dev] [PATCH 00/10] Use DP_PACKET_BATCH_FOR_EACH macro.

2017-09-20 Thread Bodireddy, Bhanuprakash
Hi Darrell, >You have many instances where you want to use >DP_PACKET_BATCH_FOR_EACH You have another series partially about this: >https://patchwork.ozlabs.org/patch/813007/ > >Also, this series mixes in other changes like creating new variables for >clarity, I >guess, and removing unneeded

Re: [ovs-dev] [patch v2 1/5] conntrack: Fix clang static analysis reports.

2017-09-21 Thread Bodireddy, Bhanuprakash
>These dead assignment warnings do not affect functionality. >In one case, a local variable could be removed and in another case, the >working pointer should be used rather than the start pointer. > >Fixes: bd5e81a0e596 ("Userspace Datapath: Add ALG infra and FTP.") >Reported-by: Bhanuprakash

Re: [ovs-dev] [patch v2 2/5] conntrack: Minor performance enhancement.

2017-09-21 Thread Bodireddy, Bhanuprakash
>Add an OVS_UNLIKELY and reorder a few variable condition checks. > >Signed-off-by: Darrell Ball >--- > lib/conntrack.c | 6 +++--- > 1 file changed, 3 insertions(+), 3 deletions(-) > >diff --git a/lib/conntrack.c b/lib/conntrack.c index 59d3c4e..c94bc27 100644 >---

Re: [ovs-dev] is there any document about how to build debian package with dpdk?

2017-09-21 Thread Bodireddy, Bhanuprakash
>we modified little code for dpdk, so we must rebuild ovs debian package with >dpdk by ourself. >so is there any guide about how to build openvswith-dpdk package? There is a guide on this here http://docs.openvswitch.org/en/latest/intro/install/debian/ - Bhanuprakash.

Re: [ovs-dev] [patch v2 3/5] conntrack: Create nat_conn_keys_insert().

2017-09-21 Thread Bodireddy, Bhanuprakash
>Create a separate function from existing code, so the code can be reused in a >subsequent patch; no change in functionality. > >Signed-off-by: Darrell Ball >--- > lib/conntrack.c | 42 +- > 1 file changed, 29 insertions(+), 13 deletions(-)

Re: [ovs-dev] [PATCH v4 5/7] timeval: Introduce time_usec().

2017-10-13 Thread Bodireddy, Bhanuprakash
>This fanction will provide monotonic time in microseconds. [BHANU] Typo here with function. > >Signed-off-by: Ilya Maximets >--- > lib/timeval.c | 22 ++ lib/timeval.h | 2 ++ > 2 files changed, 24 insertions(+) > >diff --git a/lib/timeval.c

Re: [ovs-dev] [PATCH v4 4/7] netdev: Remove useless cutlen.

2017-10-13 Thread Bodireddy, Bhanuprakash
>Cutlen already applied while processing OVS_ACTION_ATTR_OUTPUT. > >Signed-off-by: Ilya Maximets >--- > lib/netdev-bsd.c | 2 +- > lib/netdev-dpdk.c | 5 - > lib/netdev-dummy.c | 2 +- > lib/netdev-linux.c | 4 ++-- > 4 files changed, 4 insertions(+), 9 deletions(-) >

Re: [ovs-dev] [PATCH v4 0/7] Output packet batching.

2017-10-13 Thread Bodireddy, Bhanuprakash
>Hi Ilya, > >Sorry for the late response, as I was rather busy and did not find time to look >at your revisions 1 till 3. Hopefully, I can make it up looking at v4... > >I did some tests in-line with the earlier tests I did with Bhanu's patch >series. >Here is a comparison for a simple PVP test

Re: [ovs-dev] [PATCH v4 6/7] dpif-netdev: Time based output batching.

2017-10-13 Thread Bodireddy, Bhanuprakash
>This allows to collect packets from more than one RX burst and send them >together with a configurable intervals. > >'other_config:tx-flush-interval' can be used to configure time that a packet >can wait in output batch for sending. > >dpif-netdev turned to microsecond resolution for time

Re: [ovs-dev] [PATCH v4 3/7] dpif-netdev: Register packet processing cores to KA framework.

2017-09-07 Thread Bodireddy, Bhanuprakash
>Bhanuprakash Bodireddy writes: > >> This commit registers the packet processing PMD cores to keepalive >> framework. Only PMDs that have rxqs mapped will be registered and >> actively monitored by KA framework. >> >> This commit spawns a keepalive thread that

Re: [ovs-dev] [PATCH v4 0/7] Add OVS DPDK keep-alive functionality.

2017-09-07 Thread Bodireddy, Bhanuprakash
>"Bodireddy, Bhanuprakash" <bhanuprakash.bodire...@intel.com> writes: > >> Hi Aaron, >> >>>Quick comment before I do an in-depth review. >>> >>>One thing that is missing in this series is some form of documentation >>>adde

Re: [ovs-dev] [PATCH v4 2/7] Keepalive: Add initial keepalive support.

2017-09-07 Thread Bodireddy, Bhanuprakash
Hi Aaron, My reply inline. >Hi Bhanu, > >Bhanuprakash Bodireddy writes: > >> This commit introduces the initial keepalive support by adding >> 'keepalive' module and also helper and initialization functions that >> will be invoked by later commits. >> >> This

Re: [ovs-dev] [PATCH v4 0/7] Add OVS DPDK keep-alive functionality.

2017-09-06 Thread Bodireddy, Bhanuprakash
Hi Aaron, >Quick comment before I do an in-depth review. > >One thing that is missing in this series is some form of documentation added >to explain why this feature should exist (for instance, why can't the standard >posix process accounting information suffice?) and what the high-level

Re: [ovs-dev] [PATCH 12/13] conntrack: Fix dead assignment reported by clang.

2017-09-10 Thread Bodireddy, Bhanuprakash
Hi Darrell, >What version of clang are you using and in what environment ? The clang version is 3.5.0. This was seen with clang static analysis. - Bhanuprakash. > >On 9/8/17, 10:59 AM, "ovs-dev-boun...@openvswitch.org on behalf of >Bhanuprakash Bodireddy"

Re: [ovs-dev] [PATCH v3 2/4] netdev: Remove unused may_steal.

2017-09-25 Thread Bodireddy, Bhanuprakash
>Not needed anymore because 'may_steal' already handled on dpif-netdev >layer and always true; LGTM. 'may_steal' is still used by QoS policer in netdev layer. I am not familiar with Policer functionality but Just wondering may_steal isn't needed with this change. - Bhanuprakash. >

Re: [ovs-dev] [PATCH v3 3/4] netdev-dpdk: Remove useless cutlen.

2017-09-25 Thread Bodireddy, Bhanuprakash
>Cutlen already applied while processing OVS_ACTION_ATTR_OUTPUT. > >Signed-off-by: Ilya Maximets LGTM, The below redundant calls can be removed as packet cutlen is already applied in dpif layer. -Bhanuprakash. >--- > lib/netdev-dpdk.c | 5 - > 1 file changed, 5

Re: [ovs-dev] [PATCH v3 1/4] dpif-netdev: Output packet batching.

2017-09-25 Thread Bodireddy, Bhanuprakash
Hi Ilya, This series needs to be rebased. Few comments below. >While processing incoming batch of packets they are scattered across many >per-flow batches and sent separately. > >This becomes an issue while using more than a few flows. > >For example if we have balanced-tcp OvS bonding with 2

  1   2   >