Age | Commit message (Collapse) | Author | Files | Lines |
|
CSIT-1164 Create VM vhost tests for 2-node topology
CSIT-1173 Implement VM vhost KWs for 2-node topology
- Cleanup QemuUtils.py Library (PEP8, some TODOs, readability, multi queue,
ring size, qemu version detection)
- Cleanup VHOST KWs perf (Unify running QEMU on N-node topology)
- Add ability to configure Queues, Hyperthreading, automatically compute
number of CPU for VM/Testpmd and RXQ for Testpmd workload.
- Partial cleanup of configuration KWs (l2xc, l2bd) for 2-node
- Create IPv4 2-node KW
Change-Id: I8aae1355bafa651de715cd438fe706c443ea1d88
Signed-off-by: Peter Mikus <pmikus@cisco.com>
|
|
+ DUTSetup.py:424 Else clause on loop without a break statement
+ InterfaceUtil.py:400 Else clause on loop without a break statement
+ QemuUtils.py:564 Wrong continued indentation
+ SetupDPDKTest.py: Locally enabling broad-except
+ VatExecutor.py: Catching too general exception Exception
+ ssh.py:95 No exception type(s) specified.
+ HTTPRequest.py: Tolerate HTTPCodes.OK
+ multiple: Drop ":returns: None" from docstrings.
There are still several warnings present:
- R0902(too-many-instance-attributes)
- R0912(too-many-branches)
- R0913(too-many-arguments)
- R0914(too-many-locals)
- R0915(too-many-statements)
- R0401(cyclic-import)
And there are multiple blocks of similar lines,
mainly across various Setup*Test.py files:
- R0801(duplicate-code)
Change-Id: I582575cb52b85d69d268e6374852f6e74bb71052
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
+ Docstring warnings fixed.
+ Multiline param descriptions indented by 4 spaces.
- Except the PacketVerifier.py one
- I have tried several quote-like blocks, nothing works.
- Rst warnings not fixed.
- How can I fix them? They refer to temporarily created files.
+ Other improvements:
+ Python lines no longer than 80 characters.
+ :return: -> :returns:
+ Notes before params.
+ :raises
+ closing colon after exception class.
+ Description is a sentence.
+ Present tense in conditional sentences.
+ Bumped copyright year in edited files.
Change-Id: I462c194eeecb666dc146e26858486a07c990be9b
Signed-off-by: Vratko Polak <vrpolak@cisco.com>
|
|
Currently Qemu is being installed if there is a change of qsz parameter
between tests/suites. Qemu is installed always into the same directory. This
patch changes the default behavior to install qemu over and install
pathced version to separate directory. It also disables force install.
Change-Id: I0d7493a02b026a6ae4a5ea8bacf54656de9db567
Signed-off-by: Peter Mikus <pmikus@cisco.com>
|
|
Change-Id: I114c187f007c10223502669c2ba85c63d09cd84a
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
Currently there is bug that Qemu path is not correctly composed and
cannot be properly overriden due to race condition in code. This patch
is fixing functions to properly work with path.
Change-Id: I276458d2d23c7a14441a4de4a56eff6431242ab9
Signed-off-by: Peter Mikus <pmikus@cisco.com>
|
|
if unset, arch variable will default to "x86_64"
* Note on "arm64" vs "aarch64"
debian-based uses arm64
rhel-based uses aarch64
qemu binaries of both distribs uses aarch64
dpdk uses arm64
vpp uses aarch64
python machine modules uses aarch64
=> prefer aarch64 to use the same nomenclature as vpp
* add ARCH argument to:
init_dpdk.sh, install_dpdk.sh, run_l2fwd.sh, install_tldk.sh, run_tldk.sh.
default to x86_64
converts "aarch64" if needed for dpdk naming convention
* fixed terminal end detection to allow "~]# "
add dut node arch as param to all robot set bin calls
* add --target-list flag to qemu_build.sh
defaults to x86_64-softmmu
* add arch flag to all the topology files
* topologies/available/ (and example file)
* resources/tools/virl/topologies/
* set _qemu_bin path using node['arch'] in qemu_set_node()
Change-Id: If46d88d064d213d3e4c6fc584bb8e0d4b6428cb8
Signed-off-by: Gabriel Ganne <gabriel.ganne@enea.com>
|
|
Change-Id: Ib9c243b1bd379cf33a2b7747b7f97a8533000d60
Signed-off-by: Peter Mikus <pmikus@cisco.com>
|
|
Under some circumstances on a busy server, the qga channel
between the host and hte virtual guest has some garbage
characters that cause the communications to fail.
This has been seen on Centos 7.4 on a slow or busy server.
This change sends a flush to qemu guest via the qga socket at 5
second intervals until there is a non empty response from guest
before sending a guest ping.
Change-Id: I6f21c205b289169aee9d6a4072ad4e6bafafa76f
Signed-off-by: Thomas F Herbert <therbert@redhat.com>
|
|
- replace 'Qemu Kill' KW by 'Qemu Kill All'
- use self.qemu_kill_all() in case of qemu_start(self) failure
Change-Id: I19d4c64d5daef4f00548344d45f0d4943ca4f7e8
Signed-off-by: Jan Gelety <jgelety@cisco.com>
|
|
pylint repairs
deleted duplicated method from HcPersistence.py
Change-Id: I426e6c15957c995ea10bd11ae2deac75b1cab922
Signed-off-by: jan.hradil <jan.hradil@pantheon.tech>
|
|
Change-Id: I03722afe13722941e084bc01161f7c2af30e3cb1
Signed-off-by: Matej Klotton <mklotton@cisco.com>
|
|
Update the current QEMU installation script with option to override QEMU
installation and to apply additional patches. Additional patches are
applied from qemu_patches directory and subdir for specific QEMU version by
run-parts. This means that all patches for particular version are
applied.
All keywords for build QEMU are updated.
Change-Id: I0c874a96ac828dff657ee33eb87d88a8854128ad
Signed-off-by: pmikus <pmikus@cisco.com>
|
|
Add option to set path for QEMU binary and make it configurable in perf
testaces as global variable.
Change-Id: I95f4b1ab790b576f1eca02ccdb86f7a6ced80630
Signed-off-by: pmikus <pmikus@cisco.com>
|
|
- 10ge2p1x520-eth-l2bdbasemaclrn-eth-4vhost-2vm-ndrpdrdisc
Change-Id: If65a7cf21d1dbff88f39f319877d9ac1fc5d959e
Signed-off-by: Jan Gelety <jgelety@cisco.com>
|
|
Change-Id: I2e76de710b4b758eac1d24d447a9c5938f937385
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
- replace nc to socat for comunicating with qemu QGA
- change disk type to use virtio driver
- add check to bind only network interface for testpmd
Change-Id: I19ab16cbd8fec1ec78263ccfb1309a144e5af845
Signed-off-by: Matej Klotton <mklotton@cisco.com>
|
|
Change-Id: Id13c7fd51e693a273637c80cb35a159434be4826
Signed-off-by: Tibor Frank <tifrank@cisco.com>
|
|
There is an issue that booting Nested VM is successful but
QGA guest-ping or flush returns partial response. This fix suppose
to do checking until an error or timeout occurs. In case of QGA
returns correct response it is success. In case of partial
response it will log an error and continue in checking.
Change-Id: Idb2c5e5ea54bfc1e118ff62dd15696a1797dfdd9
Signed-off-by: pmikus <pmikus@cisco.com>
|
|
There is an issue that booting Nested VM is successful but
QGA guest-ping returns partial response. This fix suppose to
do checking until an error or timeout occurs. In case of QGA
returns correct response it is success. In case of partial
response it will log an error and continue in checking.
Change-Id: I8a2469287abaa9398a9206988e579912840f20c4
Signed-off-by: pmikus <pmikus@cisco.com>
|
|
Add multi-queue support when starting qemu. Full implementation
will be following in near future.
Change-Id: I743bbde67254b26a4e4e98072bad57dde728845a
Signed-off-by: pmikus <pmikus@cisco.com>
|
|
Reading information from sysfs or proc files is randomly failing
returning empty string without any error.
This fix suppose to add the retries to read the hugepages
information multiple times and raise exception if reading failed.
Change-Id: I7c88504765c39b11c81f7a8e77e5b0e32b7225fe
Signed-off-by: pmikus <pmikus@cisco.com>
|
|
Add option to configure CFS scheduler priority for QEMU
Change-Id: I76da7f536e70a23a57cf5215ab60002f95532b19
Signed-off-by: pmikus <pmikus@cisco.com>
|
|
- Switch Qemu binary to default path and remove building qemu in performance jobs
Change-Id: I5e8bb510beb672d034699192a7c454c56f2d2581
Signed-off-by: pmikus <pmikus@cisco.com>
|
|
- no functional changes
- fixes 80+ PEP-8 violations
Change-Id: Icf414778ec40d5cb44364fa69a876f9a1870c3c7
Signed-off-by: selias <samelias@cisco.com>
|
|
- Add libraries and keywords to control Qemu VM and VMapp
Change-Id: Iaaedfe240afe394f507ccaf0cf7a77c8b863acd3
Signed-off-by: pmikus <pmikus@cisco.com>
|
|
Improve Qemu affinity setting by getting the list of Qemu thread IDs via
QMP and pin each thread PID to specific host core by taskset.
Change-Id: I1b0ee8d8425cf1f97b16d6761fff0be2fadc44a8
Signed-off-by: pmikus <pmikus@cisco.com>
|
|
If Qemu requires more Hugepages than availbale in system it fails to start.
This patch is aimed to add option to configure more hugepages needed.
Default behavior is to fail to start if amount of hugepages is lower
then required.
Change-Id: I64bd8a5ec5bfa559676035c23001b12d00b026a5
Signed-off-by: pmikus <pmikus@cisco.com>
|
|
- remove mem-prealloc parameter from qemu options
Change-Id: I5a75f53da970c4aa772f29ed49bbf25c07bf2d87
Signed-off-by: Matej Klotton <mklotton@cisco.com>
|
|
- Modify runtime optimization of qemu process in qemu library
- Add ballon=none
- Add option to set affinity
- Add mem-prealloc
- Add vhost extend options
Change-Id: I721557d3e38599d783dfe0773f0593a250f8fae7
Signed-off-by: pmikus <pmikus@cisco.com>
|
|
This is not a final fix. The final fix SHOULD send a "guest-sync"
request as per"
http://wiki.qemu.org/Features/QAPI/GuestAgent#QEMU_Guest_Agent_Protocol
However, this temporary fix will eliminate the nested VM startup
failures triggered by the nested VM's guest-agent receiving a partial
guest-ping request.
Change-Id: Ice409b7c966a527b03cb71c042b476e1771676d0
Signed-off-by: Carsten Koester <ckoester@cisco.com>
|
|
JIRA: CSIT-62
Change-Id: I0affda16af8f983895c50fbc306f45f30fbb69cd
Signed-off-by: Matus Fabian <matfabia@cisco.com>
|
|
PEP8 reformat
fix typos
docstrings reformat
Change-Id: Ic48ba4e06490630808b8e2ab1ab0b046ec7eeed7
Signed-off-by: Matej Klotton <mklotton@cisco.com>
|
|
Incorporated changes from nested VM on VIRL.
Change-Id: I76ceeb2dde635c2cf0d3a53d29bf24fed72a5437
Signed-off-by: Matus Fabian <matfabia@cisco.com>
|
|
Change-Id: I22640caa9f13be09d2a837c6332d7b86cef77efc
Signed-off-by: Matus Fabian <matfabia@cisco.com>
|
|
Change-Id: I58271aff1ef558752f00593bbad5c3317d08288d
Signed-off-by: Matus Fabian <matfabia@cisco.com>
|