[RFC] Proposal for dropping FC22 jenkins tests on master branch
by Sandro Bonazzola
Hi,
can we drop FC22 testing in jenkins now that FC23 jobs are up and running?
it will reduce jenkins load. If needed we can keep FC22 builds, just
dropping the check jobs.
Comments?
--
Sandro Bonazzola
Better technology. Faster innovation. Powered by community collaboration.
See how it works at redhat.com
8 years, 9 months
AppErrors cleanup
by Allon Mureinik
Hi all,
A recent bug [1] reported as part of the translation effort alerted me to the fact that we have a lot (and I mean a LOT - over 100 per file) of deprecated, unused keys in the various AppErrors files that serve no purpose and just take up space and waste translators time when they examine them.
To make a long story short - I've just merged a patch to remove all these useless messages, and enforce via unit tests that EVERY key there should have a corresponding constant in the EngineMessage or EngineError enums.
Many thanks to my reviewers!
I know this was an tedious patch that couldn't have been too fun to review.
-Allon
[1] https://bugzilla.redhat.com/show_bug.cgi?id=1244766
8 years, 11 months
Proposal: Hystrix for realtime command monitoring
by Roman Mohr
Hi All,
I am contributing to the engine for three months now. While I dug into the
code I
started to wonder how to visualize what the engine is actually doing.
To get better insights I added hystrix[1] to the engine. Hystrix is a
circuit
breaker library which was developed by Netflix and has one pretty
interesting
feature: Real time metrics for commands.
In combination with hystrix-dashboard[2] it allows very interesting
insights.
You can easily get an overview of commands involved in operations, their
performance and complexity. Look at [2] and the attachments in [5] and [6]
for
screenshots to get an Impression.
I want to propose to integrate hystrix permanently because from my
perspective
the results were really useful and I also had some good experiences with
hystrix
in past projects.
A first implementation can be found on gerrit[3].
# Where is it immediately useful?
During development and QA.
An example: I tested the hystrix integration on /api/vms and /api/hosts rest
endpoints and immediately saw that the number of command exectuions grew
lineary whit the number of vms and hosts. The bug reports [5] and [6] are
the
result.
# How to monitor the engine?
It is as easy as starting a hystrix-dashboard [2] with
$ git clone https://github.com/Netflix/Hystrix.git
$ cd Hystrix/hystrix-dashboard
$ ../gradlew jettyRun
and point the dashboard to
https://<customer.engine.ip>/ovirt-engine/hystrix.stream.
# Other possible benefits?
* Live metrics at customer site for admins, consultants and support.
* Historical metrics for analysis in addition to the log files.
The metrics information is directly usable in graphite [7]. Therefore it
would be
possible to collect the json stream for a certain time period and
analyze them
later like in [4]. To do that someone just has to run
curl --user admin@internal:engine
http://localhost:8080/ovirt-engine/api/hystrix.stream > hystrix.stream
for as long as necessary. The results can be analyzed later.
# Possible architectural benefits?
In addition to the live metrics we might also have use for the real hystrix
features:
* Circuit Breaker
* Bulk execution of commands
* De-dublication of commands (Caching)
* Synchronous and asynchronous execution support
* ...
Our commands do already have a lot of features, so I don't think that there
are
some quick wins, but maybe there are interesting opportunities for infra.
# Overhead?
In [5] the netflix employees describe their results regarding the overhead
of
wrapping every command into a new instance of a hystrix command.
They ran their tests on a standard 4-core Amazon EC2 server with a load of
60
request per second.
When using threadpools they measured a mean overhead of less than one
millisecond (so negligible). At the 90th percentile they measured an
overhead
of 3 ms. At the 99th percentile of about 9 ms.
When configuring the hystrix commands to use semaphores instead of
threadpools
they are even faster.
# How to integrate?
A working implementation can be found on gerrit[3]. These patch sets wrap a
hystrix command around every VdcAction, every VdcQuery and every VDSCommand.
This just required four small modifications in the code base.
# Security?
In the provided patches the hystrix-metrics-servlet is accessible at
/ovirt-engine/api/hystrix.stream. It is protected by basic auth but
accessible
for everyone who can authenticate. We should probably restrict it to admins.
# Todo?
1) We do report failed actions with return values. Hystrix expects failing
commands to throw an exception. So on the dashboard almost every command
looks
like a success. To overcome this, it would be pretty easy to throw an
exception inside the command and catch it immediately after it leaves the
hystrix wrapper.
2) Finetuning
Do we want semaphores or a thread pool. When the thread pool, what size do
we want?
3) Three unpackaged dependencies: archaius, hystrix-core, hystrix-contrib
# References
[1] https://github.com/Netflix/Hystrix
[2] https://github.com/Netflix/Hystrix/tree/master/hystrix-dashboard
[3] https://gerrit.ovirt.org/#/q/topic:hystrix
[4]
http://www.nurkiewicz.com/2015/02/storing-months-of-historical-metrics.html
[5]
https://github.com/Netflix/Hystrix/wiki/FAQ#what-is-the-processing-overhe...
[5] https://bugzilla.redhat.com/show_bug.cgi?id=1268216
[6] https://bugzilla.redhat.com/show_bug.cgi?id=1268224
[7] http://graphite.wikidot.com
8 years, 11 months
[vdsm][oVirt 4.0] trimming down vm.py, stage 1
by Francesco Romani
Hi all,
as many already know, we have a problem in the vdsm/virt/* area with the size of vm.py.
As in today's master (SHA1: ed7618102e4e3aedf859adb1084eecded8e5158d), vm.py weights like:
1082 11:28:09 fromani@musashi ~/Projects/upstream/vdsm $ ls -lh vdsm/virt/vm.py
-rw-rw-r--. 1 fromani fromani 209K Nov 24 11:15 vdsm/virt/vm.py
1083 11:30:22 fromani@musashi ~/Projects/upstream/vdsm $ wc vdsm/virt/vm.py
5202 17125 213026 vdsm/virt/vm.py
I'd like to renew my commitment to bring this monster down to a sane size.
I don't think we can make it in one go. We should tackle this monster task in stages.
We have roughly five weeks left until the end of the year.
Sounds like a good timespan for the first stage, and I'm aiming for go down under
the 4000 line mark.
So we have roughly 1200 lines to eliminate somehow.
This is the initial plan I'm thinking off, in expected increasing difficulty
level, from easies to hardest:
1. LiveMergeCleanupThread:
Chance to break things: minimal
- roughly 120 lines (10% of the goal)
- can be moved verbatim (cut/paste) elsewhere, like vdsm/virt/livemerge.py,
fixing only import paths.
2. the findDrive* family of functions:
Chance to break things: minimal/very low
- roughly 50 lines (~4% of the goal)
- can be moved with minimal changes elsewhere, like vdsm/virt/storage/????.py,
3. the getConf* faimily of functions:
Chance to break things: medium-to-high in the 3.x branch, low on 4.x because we drop backward compat.
- roughly 110 lines (~9% of the goal)
- just drop them for 4.0!
- alternatively, need to figure out a proper place, maybe a new module?
4. the Devices/VM setup bits
All scatthered through vm.py (e.g. _buildDomainXML, DeviceMapping table)
Chance to break things: medium-to-low, depends on individual patches
- all summed up, roughly 200 lines (~18% of the goal)
- need serious cleanup
- no definite plan
- mostly is cut/paste or transform Vm methods into functions
5. the getUnderlying* familiy of functions:
Chance to break things: low (maybe medium on some cases, to lean on the safe side)
- roughly 550 lines (~46% of the goal)
- easy to test (just run vms!)
- can be moved in the devices/ subpackage with some changes.
- require further work in the area to properly integrate this code in the device objects
- more work is planned in this area, including move from minidom to ElementTree
So far, I've gathered ~1030 lines, so I'm short of 170 lines to reach the goal, but I'm confident
I can find something more.
The plan now is to start from the easy parts *and* in background with #4, which requires way more
work than everything else.
Thoughts, objections and suggestions all welcome, especially from storage people, because first
items in my list affects them.
Thanks and bests,
--
Francesco Romani
RedHat Engineering Virtualization R & D
Phone: 8261328
IRC: fromani
8 years, 11 months
[oVirt 3.6 Localization Question #39] "prefixMsg,"
by Yuko Katabami
Hello all,
Here is another question:
File: UIMessages
Resource IDs:
numberValidationNumberBetweenInvalidReason
numberValidationNumberGreaterInvalidReason
numberValidationNumberLessInvalidReason
Strings:
{0} between {1} and {2}.
{0} greater than or equal to {1}.
{0} less than or equal to {1}.
Question: Could you please provide some example of those messages? Comment
section in each of those strings states "0=prefixMsg" but I would like to
know what will actually replace {0}, so that I can translate them properly.
Kind regards,
Yuko
8 years, 11 months
[vdsm] strange network test failure on FC23
by Francesco Romani
Hi,
Jenkins doesn't like my (trivial) https://gerrit.ovirt.org/#/c/49271/
which is about moving one log line (!).
The failure is
00:08:54.680 ======================================================================
00:08:54.680 FAIL: testEnablePromisc (ipwrapperTests.TestDrvinfo)
00:08:54.680 ----------------------------------------------------------------------
00:08:54.680 Traceback (most recent call last):
00:08:54.680 File "/home/jenkins/workspace/vdsm_master_check-patch-fc23-x86_64/vdsm/tests/ipwrapperTests.py", line 130, in testEnablePromisc
00:08:54.680 "Could not enable promiscuous mode.")
00:08:54.680 AssertionError: Could not enable promiscuous mode.
00:08:54.680 -------------------- >> begin captured logging << --------------------
00:08:54.680 root: DEBUG: /usr/bin/taskset --cpu-list 0-1 /usr/sbin/brctl show (cwd None)
00:08:54.680 root: DEBUG: SUCCESS: <err> = ''; <rc> = 0
00:08:54.680 root: DEBUG: /usr/bin/taskset --cpu-list 0-1 /sbin/ip link add name vdsm-HIRjJp type bridge (cwd None)
00:08:54.680 root: DEBUG: SUCCESS: <err> = ''; <rc> = 0
00:08:54.680 root: DEBUG: /usr/bin/taskset --cpu-list 0-1 /sbin/ip link set dev vdsm-HIRjJp up (cwd None)
00:08:54.680 root: DEBUG: SUCCESS: <err> = ''; <rc> = 0
00:08:54.680 root: DEBUG: /usr/bin/taskset --cpu-list 0-1 /sbin/ip link set dev vdsm-HIRjJp promisc on (cwd None)
00:08:54.680 root: DEBUG: SUCCESS: <err> = ''; <rc> = 0
00:08:54.680 --------------------- >> end captured logging << ---------------------
Here in fullest: http://jenkins.ovirt.org/job/vdsm_master_check-patch-fc23-x86_64/638/console
The command like looks OK, and can't think any reason it could fail, except startup race.
Using taskset, the ip command now takes a little longer to complete.
Maybe -just wild guessing- the code isn't properly waiting for the command to complete?
Otherwise not the slightest clue :)
Bests,
--
Francesco Romani
RedHat Engineering Virtualization R & D
Phone: 8261328
IRC: fromani
8 years, 11 months
[VDSM] Coverage reports
by Nir Soffer
Hi all,
Thanks to Edward, we have now coverage reports in jenkins.
The way to access the report on jenkins is to use this url:
http://jenkins.ovirt.org/job/vdsm_master_check-patch-fc23-x86_64/<build-number>/artifact/exported-artifacts/htmlcov/index.html
Here is an example, hopefully it will be accessible when you try:
http://jenkins.ovirt.org/job/vdsm_master_check-patch-fc23-x86_64/648/arti...
Todo:
- Easy way to access the report from gerrit
It should be easy to add a link the coverage report in the comment added
by jenkins after a build finish.
- Store the coverage reports for longer time, maybe a week?
- We have only 45% coverage instead of the minimum, 100%.
Note that coverage of 25% can mean *no* code was run by the tests.
The only code
running was the functions and class definitions. Here is an example:
http://jenkins.ovirt.org/job/vdsm_master_check-patch-fc23-x86_64/648/arti...
Modules that were never imported during the tests have 0% coverage:
http://jenkins.ovirt.org/job/vdsm_master_check-patch-fc23-x86_64/648/arti...
- coverage creates lot of useless noise in the jenkins logs, need to slicense
this output.
http://jenkins.ovirt.org/job/vdsm_master_check-patch-fc23-x86_64/648/cons...
I did not find a way to do this in nosetests, may need hacking
nose coverage plugin.
- The report includes only the tests in the tests directory.
We have additional tests in lib/vdsm/infra/* which are not
included. We should
move these to the tests directory.
- The report is using absolute paths, but we like shorter relative paths.
I don't see a way to configure nosetests or coverage to generate
relative paths.
May need hacking of the generated html/json in htmlcov.
- Add "make coverage" target for running coverage locally
- An easy way to enable coverage for the functional tests or for running
a single test module.
Can be done using nosetests --cover* options. Should be documented in
tests/README, and maybe automated using a script or Makefile.
When running locally, one would like to have the script open the report
in the browser automatically:
xdg-open tests/htmlcov/index.html
- An easy way to enable coverage when testing flows in vdsm
Petr sent a patch for enabling coverage using vdsm.conf:
https://gerrit.ovirt.org/49168
We discussed adding vdsm-coverage package that will make it easy to setup
Nir
8 years, 11 months
[ATN] [master] SSO patchset were merged
by Alon Bar-Lev
Hello,
We have merged SSO patchset into master.
These kind of deep infra changes are non trivial, we hope we reduced most of the side effects within the 171 revisions and testing.
Thanks for Ravi Nori for his great effort!
The SSO is based on OAuth2 specification, full description is available[1], it is a stable supported interface of engine.
In a nut shell, the major change is that login dialog is now handled by a separate non gwt webapp, this webapp provides authentication and authorization services to other webapps.
The immediate bonus is: no need to re-authenticate to user portal and/or admin portal, maybe soon we integrate reports.
Performance bonus: if using spnego (kerberos) there is no performance penalty (double request).
Usability bonus: support many authentication sequences we were unable to provide using the previous implementation.
Regards,
Alon Bar-Lev.
[1] http://www.ovirt.org/Features/UniformSSOSupport
8 years, 11 months