Age | Commit message (Collapse) | Author | Files | Lines |
|
Also unify the order of testbeds in documents where 3n-alt appears.
Change-Id: I571dc1a2f87f45c4039080b28cfbeeb09d8d990e
Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
|
|
Change-Id: I501d892c89c6db81e5daf946b357889f7af44631
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
Signed-off-by: Peter Mikus <pmikus@cisco.com>
Change-Id: I4e39a59a192682c4bd9ebc7fa81ac727b55cf8e4
|
|
Change-Id: I6c77f633bfcae3b342d211a6b6379acb2f1adb6a
Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
|
|
- For e810cq remove dpdk-vfio-pci L2 100k and 1M scale tests
Change-Id: Icded4fdca2720fad7f51cd072ed72fc4a1b4742d
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Also added some TODOs for future methodology improvements.
Some of them will addressed in subsequent changes soon,
but not sure yet which ones exactly.
Change-Id: Ib827371b7d7a598c5fa6ab41fb7e48c73185844f
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
Change-Id: I416c4120f2d7e05582a1f4fc529dfc86a09a7788
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
Signed-off-by: xinfeng zhao <xinfengx.zhao@intel.com>
Signed-off-by: Yulong Pei <yulong.pei@intel.com>
Change-Id: I9e499df973fd9990ab02fe7008d12e78b2ab089f
|
|
Signed-off-by: xinfeng zhao <xinfengx.zhao@intel.com>
Signed-off-by: Yulong Pei <yulong.pei@intel.com>
Change-Id: If70f910003d8208e08d4dc101ac70b33e7e4162b
|
|
- For e810cq use the same dpdk-vfio-pci tests as for xxv710
Change-Id: I5857332fa8f2256b293efd610902d30a015e716d
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Change-Id: I0a9782248fee91f3db2fb5db00c800c7e5447f32
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Change-Id: If0ebb92ded94bf2b0cf8ad0d17ac2b0bd6d165a3
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Change-Id: Ia8a81135cb0dbd3b2f8877de66a5d9ed3f6e8309
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Change-Id: I5674d40362ea958c34845c22c1b5fd090d08be83
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Signed-off-by: Peter Mikus <pmikus@cisco.com>
Change-Id: I208fc73382754d3182f6f7eedc1a46ae72ef6aa2
|
|
Change-Id: Icbb4709c84b66251438b73d109606977527c6c88
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Change-Id: I4c7f37d77a9b8bda17eb3fb0b9a5bf468f1fd853
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Signed-off-by: Peter Mikus <pmikus@cisco.com>
Change-Id: I4e3ec03a87f356c0e588011b848f8dd4ab7f30aa
|
|
Change-Id: I9789e459c3db272e8f76c25df1ced266c9f5213f
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Change-Id: Iefcd777f2e51038c4efc13ff68c777efb7ed78c9
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
The test scenario is about two private networks which communication was
protected by ipsec.
The test suites are to show performance impovement about ipsec esp flow
lookup that offloaded by NIC flow director and rss.
Verified on 3n-clx and Intel E810 NIC environment, with 64B ipsec packet
flow, performance improved ~31% with 1C2T, ~110% with 2C4T, ~250% with 4C8T.
Signed-off-by: xinfeng zhao <xinfengx.zhao@intel.com>
Signed-off-by: Yulong Pei <yulong.pei@intel.com>
Change-Id: I30aec8c5115e5a6fbef88c11d1bef2624029d1b9
|
|
Signed-off-by: Peter Mikus <pmikus@cisco.com>
Change-Id: I542a16dcc1f00b0cc2a9f1bce873b0e67d60b69e
|
|
+ Replace TCP TPUT 100b tests on xxv710 NICs with 1518b ones.
+ Keep UDP TPUT tests at 100b for now.
+ Add both 100b and 1518b TCP TPUT tests to Denverton.
+ Its x553 NIC is not affected.
+ That gives us some comparison on how packet size affects performance.
Ticket: CSIT-1830
Change-Id: I5544f7ba84d4813148bcad2f9d64634939492647
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
+ Delete nat44 pps suites.
+ Rename ip4base suites, testcases, suite tags (also tg) to -tput.
+ Also robot tags like UDP_PPS and TCP_PPS.
- Keep traffic types and traffic profile names with pps.
+ They can be renamed in a different change if needed.
+ Update job specs.
- Do not update methodology yet (avoid conflict with another Change).
Ticket: CSIT-1829
Change-Id: Ic0dac441f6e8253edce88e057ab7fad1d3814c24
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
Signed-off-by: Peter Mikus <pmikus@cisco.com>
Change-Id: Ie53e1e7474542547d73bdfe2a234e6e838bdd454
|
|
Signed-off-by: Peter Mikus <pmikus@cisco.com>
Change-Id: I9987cac1493ea75eb097a0f2123768eec1ca3d73
|
|
Signed-off-by: Peter Mikus <pmikus@cisco.com>
Change-Id: I755043bc10b2bc419daf007f023742e814c3c63b
|
|
Signed-off-by: Cui,Cheng <cheng.cui@intel.com>
Change-Id: I71e257e5f16a887a40ff83d782a9ebdee9d05c34
Signed-off-by: Cui,Cheng <cheng.cui@intel.com>
|
|
Signed-off-by: Peter Mikus <pmikus@cisco.com>
Change-Id: Iac1743da3eb51c69c67e8496ad4145dcc616341c
|
|
Change-Id: I569e49967bdebb02828cd166f7a9164afbbe9205
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Change-Id: Iaf2d123bcfd594d3e35d07c981c9f4b0597c7929
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Change-Id: I2a811d5a2e618dc339a3e40c3b8c374a5051649c
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Change-Id: I6e065cf25536197c2e414fde017390d54fb3b15c
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Signed-off-by: Viliam Luc <vluc@cisco.com>
Change-Id: Iba05311055b901ca3f5c2714bf0ea1c87525af50
|
|
Signed-off-by: Peter Mikus <pmikus@cisco.com>
Change-Id: I607ccd5d4b28dca576e40b79285bb2b31a65d76c
|
|
Signed-off-by: Peter Mikus <pmikus@cisco.com>
Change-Id: I59f9df885907c7ce316e95a65fe98e9637820d01
|
|
Change-Id: If5cb8bfd0af11c9a9d35bac8de7a9c8f8c59092b
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
- This is actually bug not a feature.
- AB to be added later
- Tested on TREX and iPerf3
Signed-off-by: Peter Mikus <pmikus@cisco.com>
Change-Id: Ib6f2d13e3b9401a9fb5759e42a8a310ee11b9d41
|
|
For TCP tests, 1518b is closer to what the real traffic would be,
but as usual we want smaller sizes to stress VPP before filling line.
CPS tests remain at 64b.
Fixes: 1daa6fdc0bae284dee1b61f34534e59b60b7526a
Change-Id: Id16056b63006bb569a600fa1d9e318fd2853a5c8
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
The implementation of GTPU offload rx is to use ip4_gtpu flow MARK action
of NIC and vpp flow REDIRECT_TO_NODE and BUFFER_ADVANCE fuction to direct
received gtpu flow to gtpu4-flow-input graph node, skipped ethernet-input,
ip4-input, ip4-lookup, ip4-local, ip4-udp-lookup normal graph node
processing.
Verified on 3n-clx and Intel E810 NIC environment, single core with 64B packet,
performance improve ~33% that compare with pure software way.
Signed-off-by: xinfeng zhao <xinfengx.zhao@intel.com>
Signed-off-by: Yulong Pei <yulong.pei@intel.com>
Change-Id: I2af4589448bdb1729e4ce206a8cf3a1239c61af8
Signed-off-by: Yulong Pei <yulong.pei@intel.com>
|
|
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: Id4d84aa7268080843b099fd7ab9851234612968b
|
|
+ DPDK matrix
Signed-off-by: Peter Mikus <pmikus@cisco.com>
Change-Id: I473265ae1ebf08e1518a6b6e2998abdc750167d1
|
|
Signed-off-by: Peter Mikus <pmikus@cisco.com>
Change-Id: I47781b9d0cf6c08d12ac36774b9093c3934d9ee6
|
|
Change-Id: I2614c5688697f4b58c523cc8abe2aa34c11730a1
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Update ansible, topologies, bash scripts and job specs.
Change-Id: I834b991f281e0b29b4af9dc37fe8028460298d83
Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
|
|
Signed-off-by: Peter Mikus <pmikus@cisco.com>
Change-Id: I0309afa8a34f18a675f82ea8e696c620f5bf444b
|
|
Change-Id: Ia314d7b33278bad6071d45ee851cedcdac1ee7c3
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
Mainly added explanations and examples related to test selection.
Sore refactor may be needed to separate general information
from specifics only applicable to per-patch performance jobs.
Change-Id: Iee10f14652bab8b0bdfea732781aa2fa00f7e48f
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
Change-Id: I3e414068f8e7f619ffec9bda490d682874fe649f
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Change-Id: Ic3d084b7a2eae714f8e1437389b6d4cc9c096708
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|