As of the parent commit, there is a generic "may_fail" option through
which we could still ignore an error. Thus, command_fail is superfluous,
and avoids some extra code.
Further, in all these cases, we should better be informed about any
error, so let's not use may_fail for these tests.
There might be various occasions where a CI step should be attempted to
be executed, despite a chance of failing. For instance, we might want
to try executing an instable Iperf test, and still mark the CI as
passing if it fails.
This commit introduces a generic "may_fail" option that will mark any CI
test as succeeded, despite having failed.
Ideally, we would print a warning in the HTML code, but this is not
possible because each step emits its own HTML code, instead of the main
loop handling this. We can improve this in the future.
The cls_physim1.py is no longer needed. Function for physim deployment has
been migrated to the Cluster class. Log analysis functionality has been
moved to the Analysis class.
Refactor core network handling in CI, harmonize code
Prior to change set, there is a highly heterogeneous situation for core
network management in CI code: some log collection is done from groovy
files (when it could be done in python, which is easier for local
development), there are multiple commands to start a CN from XML files,
the code for starting a CN is always different (no one unified way to
start "any" CN), Ping/Iperf require some support in ci_infra.yaml, ...
To help with the above, refactor the handling of core networks in the CI
code as follows:
- define all CNs in ci_infra.yaml (basically how to deploy, undeploy,
collect logs, and on which node)
- make one Python class that reads from ci_infra.yaml and that allows to
start/stop previously defined CNs (and add unit tests...)
- make one XML command for start/stop of CNs, leveraging above class
- leverage the CNs defined in ci_infra.yaml consistently in
Ping()/Iperf()
- delete old code
Furthermore, the following supporting changes/bugfixes are done:
- repair Iperf2_Unidir() and add unit test
- fix a number of invalid escape sequence errors (there are some
remaining, which I will do soon)
- various bug fixes
- remove unused XML steps
This is not really used as of now, because most IPs are hardcoded. See
one of the next commits for the use of this.
To make sure that we have the correct IP, there is some code that
verifies that a hardcoded IP and the CN IP would be the same.
Note: need two separate ci_infra.yaml entries, because
- for FlexRIC, the container name is different (is the tutorial CN)
- for RFsim5G E1, we don't have ext-dn => ping UPF
Use the core network class to configure one "end" of these tests. This
will make the tests more generic and predictable, as they are explicitly
configured through two ci_infra.yaml "entities", instead of some
implicit "EPC" IP address given in configuration.
As of this commit, no XML passes this info into the framework, so no XML
step will work. As it is many changes, the following commits will add
the necessary info one by one.
These two functions use the EPC class to potentially fill in
EPC/5GC-related information, or start a packet tracer (tshark). The only
pipeline that uses these functions is RAN-gNB-N300-Timing-Phytest-LDPC,
which does not use a core network. This functionality is not used, and
can therefore be removed.
After the parent commit, the registry might contain prefixed images.
This commit adds the functionality to pull such prefixed image. Since we
rename the pulled image (to have a consistent name, from whichever
registry we might pull), remove the prefix as well so that it can be
used with other CI functionality (remove image, use in test, ...)
Examples for renaming:
- pull from internal_registry for x86
internal_registry/oai-gnb:branchA => oai-ci/oai-gnb:branchA
- pull from internal registry for ARM with "arm_" prefix
internal_registry/oai-gnb:arm_branchA => oai-ci/oai-gnb:branchA
- pull from x86 other registry (e.g., openshift)
openshift/namespace/oai-gnb:branchA => oai-ci/oai-gnb:branchA
The only currently viable way to push ARM images to our registry is to
use a custom tag prefix. Modify the python test code to specify such a
tag, and default to "" (no prefix).
* The force_local flag was introduced earlier in the main CI script
in order to run a CI test locally. The flag was missing in the
Custom_Command handling, which was still targeting the node specified
in the test configuration file. With this commit, if force_local
is enabled the node for the Custom_Command is set to localhost.
Added ./run_locally.sh script. The script takes one argument being the testcase
that one wants to run. The testcase is run locally. This means that some testcases that
utilize hardware resources will not run. However most if not all rfsimulator testcases
should run as in CI.
To this end, the CI python scripting framework was updated, adding --local flag which changes
the script behavior as follows:
- overrides <node> and <srv_node> XML elements to 'localhost' so all commands are executed locally
- Avoid running image pull, image cleanup and workspace creation steps of the scripts:
user is responsible for that
In CI we sometimes encouter fail of the gNB/UE deployment (and fail of
the pipeline) caused by unsuccsessful initialization of the USRP N310.
Restart the container if the health check fails during initialization.
Introduce a configurable parameter (from XML) to set max number of
attempts for the container deployment.
Store logs from failed deployment attempts.
If we are pulling FlexRIC, overwrite the tag to use with what has been
provided on the command line. This is because we hardcode the tag schema
"branchname-commitId[0:8]" everywhere else, but this does not apply for
FlexRIC, and we have to overwrite this here.
- pass parameter directly instead of indirectly via class
- harmonize to svr_id to read through common code in the next commits
- pass images to express which images will be deleted
- refactor function to use "with" for cls_cmd, and return success (or
not)
- reuse cls_containerize.Containerize.PullImage()
- pass parameter directly instead of indirectly via class
- always read node name and use this instead of "sometimes" looking it
up from a class member. With "sometimes" I mean that if no node name
is given, it will take it from self.eNBIPAddress, which is confusing
at best if we can give the node name directly
- consistently put the node name in the xml file
- pass parameter directly instead of indirectly via class
- call it simply "images", to harmonize with clean-function later
- harmonize to svr_id as in the previous commit
- rewrite function to use with-statement with cls_cmd
- optionally pass a tag to use (which will be used in a test, and can be
used to pull flexric)
- Introduce Pull_Image() which does the actual pulling, can pull from
any registry, and will be reused in the case of pulling from cluster.
- pass parameter directly instead of indirectly via class
- harmonize to svr_id to harmonize parameter with parameters in the next
commit (and towards a future harmonization of directly taking node
names instead of "IDs")
CI: Cleanup of cls_physim.py
- Remove unused code and file
- Rework functions for CUDA and T2 log analysis
- Check if test was successful for nr_ulsim and nr_dlsim in
RAN-gNB-N300-Timing-Phytest-LDPC pipeline
Remove CheckAndAddRoute(), as it is complicated to read, uses a
multitude of hostname/IP/password combinations. Replace with scripts on
the machines, which, as of this commit, are:
- porcepix: ~/after-reboot.sh
- ofqot: ~/after-reboot.sh
- nepes: ~/after-reboot.sh
- obelix: ~/how-to-resetup-obelix-out-of-reboot.sh
- Introduce "Run_Physim" function for running ldpctest (with CUDA) and
nr_dlsim/nr_ulsim (with T2) in RAN-gNB-N300-Timing-Phytest-LDPC test
pipeline.
- Create "cls_analysis.py" and introduce "analyze_physim" for
log analysis of physim tests.
- Define patterns to be searched in the physim logs.
- Extract and print information about total gNB RX/TX PHY processing time
from nr_ulsim/nr_dlsim logs.
- Remove "cls_physim.py" as it is not used anymore, cleanup "main.py".
pyflakes warns about various errors, addressed here. For cls_cluster.py,
remove (unknown) sys.exit and replace through Exception, as sys.exit()
has the unwanted effect that the CI script would stop. Instead, raise
an Exception that would show the problem in the HTML, and ensures that
the script runs until the end.
These are the warnings flagged by pyflakes.
cls_cluster.py:297:4: undefined name 'sys'
cls_containerize.py:39:1: 'subprocess' imported but unused
cls_containerize.py:41:1: 'threading' imported but unused
cls_containerize.py:43:1: 'multiprocessing.Process' imported but unused
cls_containerize.py:43:1: 'multiprocessing.Lock' imported but unused
cls_containerize.py:43:1: 'multiprocessing.SimpleQueue' imported but unused
cls_containerize.py:49:1: 'cls_cluster as OC' imported but unused
cls_containerize.py:50:1: redefinition of unused 'cls_cmd' from line 42
cls_module.py:28:1: 'os' imported but unused
cls_module.py:29:1: 'sys' imported but unused
cls_module.py:34:1: 'subprocess' imported but unused
cls_module.py:35:1: 'datetime.datetime' imported but unused
cls_module.py:211:11: undefined name 'sshconnection'
cls_oaicitest.py:37:1: 'pexpect' imported but unused
cls_oaicitest.py:47:1: 'cls_cluster as OC' imported but unused
cls_oaicitest.py:54:1: 'matplotlib.pyplot as plt' imported but unused
cls_oaicitest.py:55:1: 'numpy as np' imported but unused
cls_oai_html.py:34:1: 'sys' imported but unused
cls_oai_html.py:41:1: 'multiprocessing.Process' imported but unused
cls_oai_html.py:41:1: 'multiprocessing.Lock' imported but unused
cls_oai_html.py:41:1: 'multiprocessing.SimpleQueue' imported but unused
cls_physim.py:37:1: 'multiprocessing.SimpleQueue' imported but unused
cls_static_code_analysis.py:39:1: 'time' imported but unused
cls_static_code_analysis.py:40:1: 'multiprocessing.Process' imported but unused
cls_static_code_analysis.py:40:1: 'multiprocessing.Lock' imported but unused
cls_static_code_analysis.py:40:1: 'multiprocessing.SimpleQueue' imported but unused
epc.py:41:1: 'multiprocessing.Process' imported but unused
epc.py:41:1: 'multiprocessing.Lock' imported but unused
epc.py:41:1: 'multiprocessing.SimpleQueue' imported but unused
main.py:60:1: 'pexpect' imported but unused
main.py:66:1: 'datetime' imported but unused
main.py:68:1: redefinition of unused 'subprocess' from line 63
main.py:70:1: 'multiprocessing.Process' imported but unused
main.py:70:1: 'multiprocessing.Lock' imported but unused
main.py:70:1: 'multiprocessing.SimpleQueue' imported but unused
provideUniqueImageTag.py:2:1: 'os' imported but unused
This reimplements the CI deployment functions, and makes them a bit more
robust. Concretely:
- Introduce a generic "deployment tag" option that can be used to
override the tag to use for specific images. By default, it is the
current "branch-commitID[0:8]" tag name (requires the change from
function ImageTagToUse() to CreateTag(), impacting also pull/push
image functions)
- Avoid sed for image names, and use an .env file that docker-compose
picks up automatically; the deployment analyzes a potentially existing
.env file and updates instead of simply overriding. For instance, some
pipelines might use -asan images for the gNB and "normal" (non-asan)
images for UE, and a simple overwriting might make the -asan image
name tag disappear for the gNB, resulting in deployment failures).
Finally, undeployment removes the .env file, so that there are no
modifications in the repository after undeployment.
- Redo the previous behavior of autodetecting asan, and use that (the
current function always uses asan, no matter what)
- Remove deployKind/displayedNewTags globals, as they are not necessary
- Make the usedImage output in HTML slimmer
- On undeployment, print service names next to undeploy analysis, and
return success/failure.
- Make the functions generally less verbose and easier to read
Note that as of and only in this commit, deployment does not work, as
all the YAML files have not been updated to work with this updated
version. The next commit adds tests for the new deployment, and updates
one YAML file (also being used in the tests). The follow-up commit then
modifies all YAML files.
The BuildeNB steps either does things that are already implemented
(created workspace), or things that we don't use (separate build_wait),
and is generally quite convoluted when all it does is calling build_oai.
Reimplement by reusing Create_Workspace, and make build_oai in a single
step. To clean up, make a new module in which this function does not
need a class instantiation. Instead, we give all necessary parameters
directly.
Use the previous commit on returning success from each task execution
function, and implement handling of errors in the main loop. Add a
script with which this is visible using Custom_Commands.
For reasons I don't understand, the XML option reading and actual test
case execution are at different places in the code. This is problematic,
because what this led to is that options are not passed as function
arguments, but indirectly through class members. This led to a lot of
"indirect code"; it's a total mess
To improve this, move them together, so that parameters can be passed
directly as function arguments later on.
Because we reference the globals, mark them inside
ExecuteActionWithParam(). That is a Python antipattern, but necessary at
this moment. I believe that we can remove the following globals over
time:
- SSH (with new cls_cmd, which we don't need to pass)
- EPC (replaced through a class similar to ue_module, i.e., refer to
core networks on specific hosts through an ID, instead of a global)
- HTML (should use templating, where individual functions give back a
dictionary of what they want to print, and a final pass assembles the
entire HTML file)
- RAN/SCA/PHYSIM/CLUSTER/ldpc (can be replaced with stateless scripts
build/run eNB/gNB on a specific node)
Robert Schmidt