Age | Commit message (Collapse) | Author | Files | Lines |
|
The previous 60 Mpps stil leads to ~30% duration stretching.
+ Add comment on why 36 Mpps was chosen as the new limit.
Change-Id: Ic11e8ece03939bdc8680cd7bc4122373583a2f17
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
- Follow up https://jira.fd.io/browse/VPP-1934
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: Id0a26c5f67f229480332530a8531401d954f4422
|
|
- some tests need to reduce rate for ramp-up phase
- some tests need to extend trail duration in ramp-up phase
- removed 2n1l-10ge2p1x710-ethip4udp-nat44det-h1-p63-s63 suite
as nat out ports are randomly selected from available port range
so T-Rex stateless is not able to provide required out2in traffic
Change-Id: I1145496610d202f81d911e68aa819844d7600918
Signed-off-by: Jan Gelety <jgelety@cisco.com>
|
|
avf tests works, but vfio-pci tests are following, the interfaces must
be in down state (ideally unbind, as vpp cannot pick them).
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: I77af85ec4239059a5455ef68683ca129548bd7bd
|
|
- From 2n-zn2 testing
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: I9b1f0916f0f1d90a223918cfe48409d29f2ee773
|
|
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: I2ff6526b5209dabf70c6593f60162d9b8c147014
|
|
Fix 1/3: Explicitly put PF interface up (this patch)
Fix 2/3: Done on TB
Fix 3/3: VPP bug displaying VF up if underlying PF is not up
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: I45d66986ec76e6e14eebaad6828ef72724c626ab
|
|
- https://gerrit.fd.io/r/c/vpp/+/28829
- https://gerrit.fd.io/r/c/vpp/+/28896
- ip6 to be added later
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: I7b03b8ace5c5c7582d4a2f66bfd788a3ec80a916
|
|
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: I8d6108af943d729fecbcfe4867ea820a69b4eb1e
|
|
In VPP 20.05, vpp added async crypto engine that support to use QAT hardware
to do encryption and decryption, vnet/ipsec enabled async mode to use async
crypto engine.
Current async crypto engine also use dpdk_cryptodev as async handlers, in the
future it may add other native QAT driver as async handlers.
Note that async crypto engine is to support vnet/ipsec, it is different
with current existing dpdk backend which itself has ESP implementation
in plugins/dpdk/ipsec.
Change-Id: I4e6eaa7ca1eddb8b1c45212de0684fb26907119b
Signed-off-by: Yulong Pei <yulong.pei@intel.com>
|
|
Jira: CSIT-1755
Change-Id: I34baa22a49f44da3fa80d91fa2f4132c982fe610
Signed-off-by: Jan Gelety <jgelety@cisco.com>
|
|
+ Unifying code structures
- To easily plug another DUT
+ New PCI PassThrough templates
+ Improved perf stat on cores allocated in test.
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: I325f17b977314f93cb91818feddfddf3e607eb8a
|
|
+ DPDK 20.08
+ Migrate make -> meson
+ Fix all trending issues
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: I31dcb22627c0f8d17ec63c5b138a2da958b006f4
|
|
+ Bump T-Rex version. We need new features for ASTF test.
+ Apply core pining. Results in a more stable performance.
+ Tweak the number of T-Rex workers.
+ We need an even value to achieve ymmetric performance with pinning.
+ Value 8 was selected as a best compromise.
This is a combination of 3 commits.
This is the 1st commit message:
T-Rex: 2.82
This is the commit message #2:
Change Trex to CORE_MASK_PIN mode to improve performance
https://trex-tgn.cisco.com/trex/doc/trex_stateless.html#_core_masking_per_interface
Above link have below explaination,
"When the profile is symmetric, performance can be improved by pinning half of
the cores to port 0, and half of the cores to port 1, thus avoiding cache
trashing and bouncing."
The reason to change this is that to run CSIT with 100G NIC often failed with
"TRex stateless runtime error timeout", it caused by Trex can not send enough
traffic within the fixed duration.
by change to CORE_MASK_PIN mode fix the issue.
Not editing ASTF, as that supports different options.
This is the commit message #3:
Experiment: Vary number of TRex workers
With CORE_MASK_PIN, we can get more predictable time distribution.
Decided to use 8 workers, that gives good results
both for high end (RDMA-core l2patch) and low end (vhost) tests.
Change-Id: I5c61127799e0624464e960fcb980ad1c4058e744
Signed-off-by: pmikus <pmikus@cisco.com>
Signed-off-by: Yulong Pei <yulong.pei@intel.com>
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: I2781e85f44acffb4f8d7f02326ba2ca668dad0c5
|
|
- align CSIT code with VPP code changes for NAT44 deterministic
(DET44) feature
- align test names according to snat44ed tests
- remove obsolete 3-node nat tests
- remove 2n1l-10ge2p1x710-ethip4udp-snat44det-h1048576-p63-s66060288 tests
(not enough memory for such high number of sessions)
Change-Id: I9a22b99b4cfa56d18e9c7ef9c58296e202567d42
Signed-off-by: Jan Gelety <jgelety@cisco.com>
|
|
The old-style variables contain None in some places,
mainly in SRv6 proxy tests.
Change-Id: If3887a7dba051454c504b345a6a316d5d69d0139
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
Change-Id: I74641cc89d2f25d50b67d51bf2567082b420aabb
Signed-off-by: Jan Gelety <jgelety@cisco.com>
|
|
for Intel E810CQ 100G NIC, kernel driver of PF is ice, kernel driver of VF
is iavf, its VF hardware support VPP native avf driver.
Signed-off-by: Yulong Pei <yulong.pei@intel.com>
Change-Id: Ic8d86e5ee00057bbbcd09df619a38bd1371c8fd7
|
|
- continuation of https://gerrit.fd.io/r/c/csit/+/26898 as there was
reached limit of changes (1000)
Jira: CSIT-1711
- udp synthetic profiles w/o data packets
- udp cps perf tests, phase I (no special "search cps" KW)
Part I means that we are using MRR tests to collect traffic data
until there is ready new CPS test type with corresponding algorithm.
Change-Id: I0d30feb9ecf1d0bff937152656f8eb422f831378
Signed-off-by: Jan Gelety <jgelety@cisco.com>
|
|
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: I20098fca8fb513accef3edc9a72bfd3c56bf9be2
|
|
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: I1b535ea61ab68f6e37989ffc942979cdfd24f55e
|
|
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: Iab84aff31a23bb9d8e1165f5314004803fd8a501
|
|
+ Measure latency in 90/50/10/0% PDR loads in ndrpdr tests.
+ Do not measure latency anywhere else.
- Needs manual editing to re-enable in soak tests.
Change-Id: I69fa11bfcf71012f683061c5effea52a1be91620
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
To avoid runs such as:
https://jenkins.fd.io/job/vpp-csit-verify-perf-master-3n-hsw/340/
Change-Id: I1b30d5f440ddf8ff32b11265b2ac2176f4b9a665
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
- provide base routines to run T-Rex in advanced stateful mode
Change-Id: Ib0dc5f2919c370753335f6446860683dc4b12d93
Signed-off-by: Jan Gelety <jgelety@cisco.com>
|
|
Change-Id: I3bbe1fe0073ddeead5219993675f24955e8c3dfd
Signed-off-by: Peter Mikus <pmikus@cisco.com>
|
|
- test/suite/global
- binary logic is not working
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: Ia3d81cbf2c5f04d1093a0a408c84a9ffc6f3eef0
|
|
Change-Id: I31c2d7744b5cd3021132fb188480b8edec74986c
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
+ Latency measurements need more than 9000 pps.
- Previously 0% measurement used 9500 pps.
Change-Id: Ic0841de096dfa8a61329f98aa1ba6d3f0ce60c66
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
Change-Id: If0975b1d54882390c5be418927e2961d0f4c8429
Signed-off-by: Jan Gelety <jgelety@cisco.com>
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
VPP uses MAKE_PARALLEL_FLAGS or MAKE_PARALLEL_JOBS to limit the number
of cpus to use during build, so emit a line on stdout if it's used.
Change-Id: I669398d474d172abb6c848a45f24f1bdd56990d8
Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
|
|
Change-Id: If61783fb717757c6189f06924412bd079e15a08f
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Change-Id: I6da359d25edc415e44263d3f85f166369e564987
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
Change-Id: I636f020e97df1b37ac8b6a30af511eebe611b56f
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
Previously, number of directions was not taken into account.
Also, ideally PLRsearch never reports value under the hard minimum,
so successful results are now required to be more than 10% better.
Change-Id: I8622726b97bd1da3e139c8044a2932837fc268b7
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: Ie88f0df239725a4de62d727e1923cdb3ad040809
|
|
for example, below will trigger VPP with default code path test in CSIT,
"csit-2n-clx-perftest mrrAND64bANDnic_intel-xxv710ANDeth-l2bdbasemaclrnNOTdrv_avf"
and below will trigger VPP with avx512 code path test in CSIT,
"csit-2n-clx-perftest icl mrrAND64bANDnic_intel-xxv710ANDeth-l2bdbasemaclrnNOTdrv_avf"
Change-Id: Id3abf932203ce00f34f8643c5bc4d5fc4d3e6efb
Signed-off-by: Yulong Pei <yulong.pei@intel.com>
|
|
This reverts commit 925c5bfcffb2bc6b833a8e64b7f3a06bbca52197.
It was affecting 2005 performance, confusing comparisons.
The revert includes the RDMA limit.
Change-Id: I9c31b061ee59bc5cb14d623ac98a65763ef75f1c
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
- Create the same directory structure for report-iterative as it is
for report-coverage.
- Move files from report-iterative directory to created sub-directories.
- Rename the moved files to keep the convention defined in
report-coverage.
- Update the corresponding functionality in
resources/libraries/bash/function/common.sh
The patches:
- https://gerrit.fd.io/r/c/csit/+/27486
- https://gerrit.fd.io/r/c/csit/+/27484
- https://gerrit.fd.io/r/c/csit/+/27485
must be updated.
Change-Id: I1aaf22690e3df5abf56893bb1d2d380debc54ddd
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Signed-off-by: Maros Mullner <maros.mullner@pantheon.tech>
Change-Id: Ib5f58f60a1409ed139e2846793bf52fdc02a6571
|
|
The test names are unique without it and the information doesn't add
anything extra.
Change-Id: Idc7d6d1d21c8c05691e1757227a0a3787406d370
Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
|
|
ARP not resolved as ARP plugin is not in use.
Signed-off-by: pmikus <pmikus@cisco.com>
Change-Id: I655d39794f99e32a2f3905154d78e83853356ccf
|
|
Directive #include <file> in GCC implementation searches for files
in a standard list of system directories, which leads to a build
error on Taishan arm machines:
/tmp/openvpp-testing/dpdk/lib/librte_ethdev/rte_ethdev.h:4287:10:
fatal error: rte_ethdev_core.h:
No such file or directory #include <rte_ethdev_core.h>
Fix this by replacing the directive with #include "file" which searches
for files first in the directory containing the current file and other
directories only after that.
Change-Id: If283b81c992c001a2f980312cdd99053c3d703d9
Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
|
|
Change-Id: I12f21a4a7ae7c8593c42fa8e6cc4d65e36d41270
Signed-off-by: Jan Gelety <jgelety@cisco.com>
|
|
There could be directories under /dev/hugepages/ so remove those as
well.
Change-Id: I8ecdde159b8e0049460d0bcdc8b1a2c85b1ed25f
Signed-off-by: Juraj Linkeš <juraj.linkes@pantheon.tech>
|
|
Change-Id: I1159fb29d43f61628922cdbd51313f710aa72f4a
Signed-off-by: Jan Gelety <jgelety@cisco.com>
|
|
- split job specs to directories
Change-Id: I9a71621216ab0bdf85d147ce03f473e7cbb0b266
Signed-off-by: Jan Gelety <jgelety@cisco.com>
|
|
- note: currently dpdk report-coverge test sets are equal to dpdk
report-iterative test sets
Change-Id: Ib9a718bfea4b24842742f49ee2eca022aa8a08e5
Signed-off-by: Jan Gelety <jgelety@cisco.com>
|
|
- job will be used to collect vpp performance iterative data
for csit report
Change-Id: I4cff5caa0a97d08b11f3eb9b72963deaf1b04c73
Signed-off-by: Jan Gelety <jgelety@cisco.com>
|