- Remove usless abstraction of db.api
- Split API methods to schema management and just record related ops
- Remove part of oslo.db and oslo.utils dependency (need one more patch)
- Improve performance & refactor mulitple requests
Change-Id: I56ca311ea354409e9b19511dca77cda27be5d52b
* fix lable of database section. it should be [database] instead of [db]
* create `/home/rally/data` dir by rally user instead of root, so rally
could operate the database there.
* modify the volume path
* update the docs
Change-Id: Ia4df80e2ce570ede4582c875c0550d5ba8da60d9
ChenHb and Hai Shi were added long time ago to rally core team, but the
info page was not updated yet.
Change-Id: I59f7d7e08127a99c3d8447aa15f04549d15a07df
- Remove translations
Nobody is using translations for Rally and I don't think that
anybody is going to use it. Target auditory for Rally are
developers/operators which usually know well english.
For me this looks like waste of resources, performance
degradation (cause we are calling _()), complexity
(+1 thing that you need to know)
- Pass to log already formatted strings
It's very bad because in case of wrong formatting, it
doesn't fail instead just writes errors to the logs,
as well information about trace is lost, so it's super
hard to fix it
Log wrapper doesn't allow to use LOG anymore for
formatting strings
All places are fixed
- Improve logging of exceptions
LOG.exception() already logs exception, which means it's bad idea to
pass str(e) to it. Instead we should provide clear description of what
happend. Improved few places to write warnings or exceptions in case
of different level of logs. In few places just use LOG.exception
- Part of log messages were improved and simplified
Depends-On: If23d874e8b73de12ba2b8c4e028a55543af6381b
Change-Id: Ibc1e1f4f554649d14b8fe4801557b83922ecefe3
In case if we specify only argument user_choice_method schema
will match both oneOf sections and fail, however it's right schema
and should pass (in other words we need anyOf)
As well fix doc extension to work with anyOf as well
Change-Id: I2b2a3a1c42277a83511cdac81496b5240d983d33
Every day Rally step by step closer to the point where Rally is going to be
generic framework and it would be possible to split Rally & Rally OpenStack Plugins
One of the blockers for doing this is certification & samples directories and
their contents. Current apporach creates a lot of confusion:
- Rally Users are running samples against production
- Rally samples are actually not that usefull, because creating from them
tasks takes a lot of time
- Rally Users don't know what certification task is and how to use it
- Samples are hard to ship/find in packaged way (when rally is installed)
- Samples are going to be impossible to ship after the split
- We have to keep hunderds of yaml/json files in actual state
in our repo (extra work for developers)
Proposed approach is to:
- Create mechansim CLI command that generates samples using plugin names
- Rename certification -> tasks
- Put in tasks directory pre created and tested tasks for different use cases:
key performance metrics, functional testing, load testing of key functionality,
ha Testing
- Remove samples as they are not needed anymore
Change-Id: I4c45224c3af637d17dab0edea038c69a2bae8b1a
Adhering to coding conventions. Refer to ``Code conventions`` at
https://docs.openstack.org/contributor-guide/ for details.
"." is synonymous with source in bash, but not in POSIX sh, so we
should use "." if the script is run by "/bin/sh".
"/bin/sh" is often used in cron.
Change-Id: I0ee691d84a19f3f06f111d50e23c0c9c80ebd402
Having 2 entrypoint seems redudant and complicated for end users,
so instead of rally-manage db <commands> let's use rally db <commands>
This reduces amount of code and commands needed to know to work with
Rally which is actually good.
This patch doesn't remove rally-manage just deprecates it
Change-Id: Ib35db2d22174cd3d3eee0d87358ad07a91dce0d9
1) Update doc links according to OpenStack document migration
2) Use https instead of http for docs links
Change-Id: Id606f43f486e98fc33b699b36e468de954ec677d
In a lot of placeses we are using word "benchmark" which
can mean workload, subtask, or test case which is very confusing.
This patch partially address wrong usage of "benchamrk" word
Change-Id: Id3b2b7ae841a5243684c12cc51c96f005dbe7544
Last time I removed step 9 tutorial for deploying openstack from
rally document (https://review.openstack.org/#/c/475932/).
However, I forgot to update the tutorial step number after that.
This just updates the tutorial step number to make it continuous.
Change-Id: Ib16408350b147eeb70b5de9b18f8014ac12714a6
