Home - dianaclarke/openstack-notes GitHub Wiki

OpenStack Journal

Week of April 17, 2017
- Slow week, vacation
- Tripleo quickstart again - notes
- Post tripleo quickstart - notes
- Test - Stop using CentOS buildlogs CDN
- Review - Add novaclient client_name and client_version to user-agent
- Review - Tell people that the nova-cells man page is for cells v1
- Ah, good to know: "set -e causes the shell to exit if any subcommand or pipeline returns a non-zero status"
- Review - Fix dlrn link to point to trunk.rdoproject.org
- Red Hat Bug #1426284 - [Tracker] API Version Issue: Keystone V3, Glancev2.3 and Cinder V3
  - openstack endpoint list http://paste.openstack.org/show/607382/
  - keystone: https://192.168.24.2:13000/v2.0
  - glance: https://192.168.24.2:13292
  - /etc/nova/nova.conf http://paste.openstack.org/show/607383/
  - [keystone_authtoken]/auth_uri=https://192.168.24.2:13000/v3
  - [glance]/api_servers=http://192.168.24.1:9292
- Red Hat Bug #1441962 - Remove default option for keymap in code and config files
Week of April 10, 2017
- Slow week, vacation
- Tripleo quickstart again... snapshot appears to be b0rked - notes
- These upgrade notes might help: https://gitlab.cee.redhat.com/mandreou/OSP10-OSP11-Upgrade
- Red Hat Bug #1434931 - Fix endpoints version used in nova.conf
  - Red Hat Bug #1426284 - [Tracker] API Version Issue: Keystone V3, Glancev2.3 and Cinder V3
  - From my newton oooq nova conf: [keystone_authtoken] auth_uri=https://192.168.24.2:13000/v3
- Red Hat Bug #1440823 - No comments in nova.conf in the overcloud
- Did the Designate team discuss the broken gate in today's meeting?
  - http://eavesdrop.openstack.org/meetings/designate/2017/designate.2017-04-12-17.00.log.html
  - "mugsie and I still need to work on fixing the gates"
- Read: https://www.rdoproject.org/documentation/rdo-packaging/#packaging-overview
- Read - high level design doc for tripleo quickstart
Week of April 3, 2017
- Slow week, sick + vacation
- TODO: Ansible tutorials
- Try a tripleo upgrade again
- Review downstream ansible upgrade job
- Patch - Add tests for compute v2.40 microversion
- More broken designate gate work, working with the eventlet folks here
- Review - Disable using upper-constraint
- Review - WIP Install upstream patch of eventlet
- [RFE] add snapshotting functionality
- Red Hat Bug #1434931 - Fix endpoints version used in nova.conf
  - Red Hat Bug #1426284 - [Tracker] API Version Issue: Keystone V3, Glancev2.3 and Cinder V3
Week of March 27, 2017
- Red Hat Bug #1435747 - Incorrect ownership of /var/log/nova/nova-placement-api.log results in scheduler failing to return valid hosts
  - https://review.openstack.org/#/q/topic:tripleo/nova/placement
  - Perhaps something like (via eglynn): https://github.com/rdo-packages/keystone-distgit/blob/047e9ec45fa69a828a1f6b274fbbaf2bd7b40a69/openstack-keystone.spec#L232
  - undercloud, post upgrade, CI: http://paste.openstack.org/show/604488/
  - overcloud, post upgrade, CI: http://paste.openstack.org/show/604498/
- Ah, so this is what they mean when they say do it in the spec file: https://github.com/rdo-packages/nova-distgit/blob/0ef48b159dccc2573ce52199132bab495970708e/openstack-nova.spec#L799
- Try a tripleo upgrade from Newton to Ocata for the first time - notes
- Patch - typo: centos7-oacata -> centos7-ocata
- Patch - Enable posargs for the py35 tests
- Patch - Python 3 - replace urlparse with six.moves.urllib
- Patch - Blacklist newly failing python 3 tests
- Patch - Fix python-novaclient deprecation warning
- Patch - Fix 'nova list --fields' error in no instances case
- Good to know: auditctl http://pastebin.test.redhat.com/469408
- Review - Explicitly specify arguments of server_groups creation
- Review - Fixing a potential unbound local error
- Read - https://docs.openstack.org/developer/requirements/
- Revisit - https://pip.pypa.io/en/stable/user_guide/#constraints-files
- Pretty certain the designate gate failures are b/c of the u-c bump from eventlet==0.19.0 to eventlet==0.20.1
- Shadow - Red Hat Bug #1436769 - Nova API is very slow on an IPV6 environment with Nova control plane services running on a custom role
- Review - placement: fix the apache virtual host
- Review - Add SOA to RECORD_TYPES in sqlalchemy.tables
  - http://logs.openstack.org/64/451264/1/gate/gate-shade-dsvm-functional-neutron/6a7432a/logs/screen-designate-api.txt.gz?level=ERROR
  - http://docs.sqlalchemy.org/en/latest/changelog/migration_11.html#the-enum-type-now-does-in-python-validation-of-values
- Red Hat Bug #1437057 - openstack network list gives Unexpected API Error
  - devstack, and undercloud output: http://paste.openstack.org/show/604848/
  - example success wih --debug to compare against: http://paste.openstack.org/show/604850/
  - overcloud success example: http://paste.openstack.org/show/604856/
  - service catalog that works: http://paste.openstack.org/show/604863/
  - openstack catalog list: http://paste.openstack.org/show/604869/ (success case)
  - my undercloud: http://paste.openstack.org/show/605064/
- Review - Add tests for compute v2.40 microversion
Week of March 20, 2017
- Review STU blurb from tech writers
- Review - Add tests for compute v2.40 microversion
- Red Hat Bug #1426693 - fence_compute: project_id changed to project_name as a parameter for nova python client
- Hmm... I can't seem to build the nova docs anymore: sphinx.errors.SphinxWarning: WARNING: dot command 'dot' cannot be run (needed for graphviz output), check the graphviz_dot setting
  - fixed: brew install graphviz
- These aren't rendering correctly, fix them (for 2.40): https://docs.openstack.org/developer/nova/api_microversion_history.html
  - Patch - Fix rest_api_version_history (2.40)
- Review - Drop deprecated aggregate-update positional args
- Bug #1671681 - Permission denied: '/var/log/nova/nova-manage.log'
- Set up a tempest dev env again, notes from last time
  - current tempest error: http://paste.openstack.org/show/603639/
  - perhaps compare my conf with another conf? http://ix.io/p95
  - try with a fresh devstack & tempest combo
  - finally got it working: http://paste.openstack.org/show/603670/
  - schema hack: http://paste.openstack.org/show/603800/
- What do the cells etc looks like in master devstack? database tables
- Research the Designate project - notes
- Red Hat Bug #1435747 - Incorrect ownership of /var/log/nova/nova-placement-api.log results in scheduler failing to return valid hosts
Week of March 13, 2017
- Slow week vacation, work on SQLAlchemy instead
- Red Hat Bug #1428572 - Nova Cannot write to /var/log/nova/nova-manage.log on Undercloud
  - Ask on #puppet-openstack about the consumers of puppet-nova: http://eavesdrop.openstack.org/irclogs/%23puppet-openstack/%23puppet-openstack.2017-03-13.log.html#t2017-03-13T12:40:43
  - WIP Patch - Run nova-manage commands as the nova user
- Review - Add tests for empty association set comparison
- Patch - add autoescape option to startswith, endswith, and contains
- Look for next tasks here: https://bitbucket.org/zzzeek/sqlalchemy/issues?status=new&status=open&milestone=1.2
- Triage - 'IS (NOT) DISTINCT FROM' operator missing
- Try working on this one: https://bitbucket.org/zzzeek/sqlalchemy/issues/3673/detect-and-raise-for-version_id-is-null
  - http://docs.sqlalchemy.org/en/latest/orm/versioning.html
  - Patch - WIP - detect and raise for version_id is NULL
  - Notes for the third (polymorphic) case: http://paste.openstack.org/show/602876/
- https://bitbucket.org/zzzeek/sqlalchemy/issues/3911/hybrid-decorators-are-missing-getter
- https://bitbucket.org/zzzeek/sqlalchemy/issues/3912/hybrid-decorators-mutator-decorators
  - from cpython: https://gist.github.com/dianaclarke/fd376343e3c7aed824505c8744b2300c
  - Patch - WIP - hybrid_property subclassing
- Answered - [openstack-dev] [codesearch] how to exclude projects?
- Review - Introduce fast8 tox target
- Review - hound: ignore deb-* projects
- Red Hat Bug #1432187 - Lack of parameter validation in nova usage REST api
  - private comment
Week of March 6, 2017
- TODO: Review these docs: https://docs.openstack.org/developer/tripleo-docs/containers_deployment/
- TODO: Read - https://dprince.github.io/docker-puppet.html
- Red Hat Bug #1427078 - No comments in nova.conf in the overcloud.
  - configure.d/, newton: http://paste.openstack.org/show/601802/
  - configure.d/, master: http://paste.openstack.org/show/601806/
  - https://github.com/openstack/dib-utils/blob/694d2988cb62e1fc2d10631b9dea6ecba6c53991/bin/dib-run-parts#L99
  - Output from dib-run-parts (via owalsh): http://paste.openstack.org/show/601808/
  - Patch - Remove nova.conf truncate/refresh workaround
- Red Hat Bug #1368588 - Permission denied: '/var/log/nova/nova-manage.log'
  - I suspect this is where that log path (/var/log/nova/nova-manage.log) comes from: https://github.com/openstack/oslo.log/blob/639404266f415de0cd805f3ae5d6c00a9e007779/oslo_log/log.py#L77
  - Why is tripleo/packstack/etc running the nova db migrations as root (rather than as the nova user)? Look at the annotations for these callers: http://codesearch.openstack.org/?q=nova-manage%20db%20sync&i=nope&files=&repos=
  - I think this is two fixes: run db sync as nova (puppet-nova, I think), and chown the nova logs dir on upgrade (ansible upgrade tasks in heat-tripleo-templates, I think)
  - nova crontab on my undercloud: https://paste.openstack.org/show/601987/
  - Patch - Run nova-manage commands as the nova user
  - Created Bug: https://bugs.launchpad.net/puppet-nova/+bug/1671681
- Review - Fix ValueError when incorrect metadata passed
- Review - Remove functional tests for removed commands
- Review - Deprecate list-secgroup command
- Read - http://hardysteven.blogspot.ca/2016/08/tripleo-deploy-artifacts-and-puppet.html
- Review - Add support for showing flavor details (v2.43)
- Review - Teach simple_tenant_usage about cells
- Read - http://rspec-puppet.com/tutorial/
- Running just one test at a time: $ bundle exec rspec spec/classes/nova_db_sync_spec.rb
- Review - Remove py34 tox env and pypi classifier
- Review - Add undercloud indicator to stackrc
- Review - Add stack name indicator to overcloudrc
- Review - overcloudrc: set OS_AUTH_TYPE
- Review - Fix typo in package_update.yaml
- Red Hat Bug #1428572 - Nova Cannot write to /var/log/nova/nova-manage.log on Undercloud
  - For bug triage, another report of the same bug I'm already working on
  - https://docs.puppet.com/puppet/latest/types/file.html
  - https://docs.puppet.com/puppet/latest/lang_relationships.html
  - "t-h-t includes profiles from puppet-tripleo which include puppet-nova manifests"
  - https://github.com/openstack/puppet-tripleo/blob/master/manifests/profile/base/nova.pp
  - https://github.com/openstack/puppet-tripleo/tree/master/manifests/profile/base/nova
- Review - Add version pin for image related function
- Review - Fix 'nova list --fields' error in no instances case
  - TODO: fix up this patch if the original author doesn't circle back
Week of February 27, 2017
- Meeting - tripleo environment setup tips
- tripleo quickstart notes, part 2
  - newton failed after 143m13.532s minutes - https://gist.github.com/dianaclarke/54ba9d943d8847c41ebdbd0e428ad745
  - master failed after 207m23.840s minutes - https://gist.github.com/dianaclarke/973476af4c145aa2d63b0388a78160a1
  - OMG! It worked! newton, 166m0.430s minutes - https://gist.github.com/dianaclarke/a2916a284003ad55ae6c6a8e7be64859
- TODO: Clean-up any downstream bugzilla issues assigned to me
  - Revisit - Red Hat Bug #1373318 - Can't re-start a stopped instance after trying to volume-update on it
- It looks like the tripleo-quickstart-extras repo isn't being indexed by http://codesearch.openstack.org/
  - This will probably point me in the right direction: https://github.com/openstack-infra/infra-specs/blob/master/specs/code-search.rst
  - Create tripleo-quickstart-extras project - https://github.com/openstack-infra/project-config/commit/b235adc63bd9f2462b86c9359244411211827b97
  - https://docs.openstack.org/infra/system-config/codesearch.html?highlight=hound#maintenance
    - https://github.com/openstack-infra/project-config/blob/master/gerrit/projects.yaml
    - https://git.openstack.org/cgit/openstack-infra/system-config/tree/modules/openstack_project/manifests/codesearch.pp
    - https://github.com/openstack-infra/jeepyb/blob/master/jeepyb/cmd/create_hound_config.py
    - When I clone jeepyb & run the create-hound-config command on the current projects.yaml, here's the resulting config.json:
      - https://gist.github.com/dianaclarke/1533448ed33232f5c1c348ab57cb884e
    - Is tripleo-quickstart-extras in there? Yup! Hmmm... is it just not up-to-date?
      - https://gist.github.com/dianaclarke/1533448ed33232f5c1c348ab57cb884e#file-config-json-L8025
    - Hmmm... tripleo-quickstart-extras is also not listed here: http://codesearch.openstack.org/excluded_files.html
    - Asked on IRC: http://eavesdrop.openstack.org/irclogs/%23openstack-infra/%23openstack-infra.2017-03-02.log.html#t2017-03-02T11:20:59
    - Created bug https://storyboard.openstack.org/#!/story/2000889 (not sure if that's the right place)
    - Fixed by the infra team (puppet was disabled)
- Review polarion test plan for upgrades
- Provide additional STU text for technical writer
- TODO: Follow along here: https://review.openstack.org/#/q/owner:%22Oliver+Walsh%22+status:open
- TODO: Review this in detail & perhaps profile it: https://review.openstack.org/#/c/343734/
- Read - https://docs.openstack.org/developer/puppet-openstack-guide/
- Red Hat Bug #1427078 - No comments in nova.conf in the overcloud.
  - Upstream Bug #1568820 - Duplicate sections in generated config
  - https://github.com/openstack/tripleo-puppet-elements/blob/6f338b7df06c800b59488abb4a8cf0f9d7c90fbd/elements/truncate-nova-config/install.d/11-truncate-nova-orc-install
  - What is tripleo-puppet-elements? "These elements are used to build disk images for deploying OpenStack via Heat."
  - What's in 40-truncate-nova-config?, http://git.openstack.org/cgit/openstack/tripleo-puppet-elements/tree/elements/truncate-nova-config/40-truncate-nova-config
  - So truncate and then os-refresh-config?, https://github.com/openstack/os-refresh-config
  - If I've understood correctly, incorrect oslo config usage in nova resulted in duplicate sections in the nova config. The underlying cause was fixed/prevented in oslo.config 3.10.0, but nova stable/mitaka (OSP9) is oslo.config>=3.7.0, and nova stable/newton (OSP10) is oslo.config>=3.14.0 (and upper-constraints is oslo.config===3.9.0 for stable/mitaka).
  - As a workaround, tripleo added a script (40-truncate-nova-config) that truncates the existing nova config and then regenerates it using os-refresh-config when it builds the disk images for OSP director deploys (which if I had to guess is where the comments are disappearing, but I'm not 100% certain yet).
  - Since nova no longer incorrectly uses oslo.config (I think) and olso.config now does the right thing when misused, I believe this workaround in tripleo-puppet-elements can be removed.
  - nova.conf does indeed not have comments: http://paste.openstack.org/show/601425/
- Red Hat Bug #1368588 - Permission denied: '/var/log/nova/nova-manage.log'
  - I messed that triage up :( I'd like to think I wouldn't do that now 6 months later, but...
  - Watch "Production Chain for Everyone" by Mike Burns
  - So the wrong turn I took was noting private comment #7 saying that they were using packstack, and then matching that fact with this other bug title: PackStack installs Nova crontab that nova user can't run. I get now that packstack can be used to install either RDO or OSP.
  - Now, has this bug actually been fix anywhere yet? puppet-nova adds the cron. Start there.
  - https://github.com/openstack/puppet-nova/commits/master/manifests/cron/archive_deleted_rows.pp
  - Who owns what in this log dir? http://paste.openstack.org/show/601369/ and http://paste.openstack.org/show/601373/
  - The first writes in that ^ file are the db migrations (nova-manage db sync, I think)
  - PackStack sets the destination to /dev/null, http://git.openstack.org/cgit/openstack/packstack/tree/packstack/puppet/modules/packstack/manifests/nova/api.pp#n62
  - Default destination in archive_deleted_rows.pp is '/var/log/nova/nova-rowsflush.log'
  - Bug is reporting /var/log/nova/nova-manage.log, go back to the logs in the bug
  - From my undercloud: undercloud_install.log:2017-03-02 03:06:55 - Notice: /Stage[main]/Nova::Cron::Archive_deleted_rows/Cron[nova-manage db archive_deleted_rows]/ensure: created
Week of February 20, 2017
- Holiday Monday
- Red Hat Bug #1421633 - Any workaround of Bug #1444841 “Resize instance fails after creating host aggregate”
  - What's a host aggregate? "Host aggregates can be regarded as a mechanism to further partition an availability zone; while availability zones are visible to users, host aggregates are only visible to administrators." (https://docs.openstack.org/developer/nova/aggregates.html)
  - What's resize under the hood again? "cold migration (with resize): Power off, move, resize" (http://www.danplanet.com/blog/2016/03/03/evacuate-in-nova-one-command-to-confuse-us-all/)
  - Upstream bug: Resize instance fails after creating host aggregate
  - Duplicate? resize failed because of instance's invalid availability zone property
  - https://blog.russellbryant.net/2013/05/21/availability-zones-and-host-aggregates-in-openstack-compute-nova/
  - Not sure what to do next... perhaps try reproducing it? (some of my old resize notes in case they help)
  - Manual testing notes: host aggregate notes
- This devstack warning is annoying, look into fixing it: http://paste.openstack.org/show/598407/
  - Here's what the default devstack, stable/ocata, openrc looks like: http://paste.openstack.org/show/599900/
  - https://github.com/openstack/python-novaclient/blob/540f9ceb461e4ad55b83e873419f46d5b75bcac8/novaclient/shell.py#L844
  - Diff between client instantiations: http://paste.openstack.org/show/600048/
  - Perhaps needs more changes similar to this: https://github.com/openstack/python-novaclient/commit/9bbe5a87b7df62a7962debba5db7c96555da6761
  - Yup, the warning goes away when I make similar changes to the second instantiation. Now to confirm that's correct though ;)
  - Patch - Fix devstack python-novaclient warning
- TODO: Perhaps try to contribute here? https://blueprints.launchpad.net/nova/+spec/cells-aware-api
- Review - Avoid limit=-1 description
- Is python-novaclient 7.1.0 in brew now? Yup :)
- TODO: Perhaps this will help: https://keithtenzer.com/2015/10/14/howto-openstack-deployment-using-tripleo-and-the-red-hat-openstack-director/
- Write up initial thoughts on trying to learn tripleo
- I think we should probably backport this too: Fix devstack python-novaclient warning - stable/ocata cherry pick
- See if there are any new downstream Nova bugs I can triage while folks are at PTG:
  - https://url.corp.redhat.com/nova-NEW-non-RFE-or-tracking-last-14-days
  - https://url.corp.redhat.com/novaclient-NEW-non-RFE-or-tracking-last-14-days
  - Nope, I'm not seeing anything dum-dum diana can help with, lol
- Can I help here? Follow along either way. Setting quota fails saying admin project is not a valid project
- Look at the commits to see why the virtual tripleo environment docs are deprecated: https://github.com/openstack/tripleo-docs
  - https://docs.openstack.org/developer/tripleo-docs/environments/virtual.html
  - It just says that they are, not why.
- Review - doc: Don't put comments inside toctree
- Review - doc: Remove literal block
- Review - Reduce time and memory usage of tenant usage.
- Try the tripleo quickstart again, this time on centos. Try not to murder anyone ;)
Week of February 13, 2017
- Prepare some tripleo questions
- TODO: How are the STU dependencies looking now for OSP11?
  - python-novaclient 7.1.0 still not in brew
  - What about RDO? https://github.com/redhat-openstack/rdoinfo/blob/0315dbd786039bc44423b7d24ad4a1695c8290c4/rdo.yml#L1334
- TODO: Revisit this patch series - Add nova service support for composable upgrades
- https://etherpad.openstack.org/p/nova-ptg-pike
- Watch more of the TripleO bootcamp videos: https://mojo.redhat.com/docs/DOC-1122111
- See if this video helps: Deploying OpenStack on OpenStack with TripleO, Heat and Ironic - Steven Hardy
- How about this video? tripleo-quickstart demonstration - John Trowbridge
- Okay, now what? - [notes]
  - http://docs.openstack.org/developer/tripleo-quickstart/accessing-undercloud.html
  - http://docs.openstack.org/developer/tripleo-quickstart/accessing-overcloud.html
- Hmmm... Simple tenant usage bug? http://logs.openstack.org/44/434344/1/check/gate-shade-dsvm-functional-libs-nv/495b14a/console.html#_2017-02-15_16_10_01_704583
  - I think this is just a bad test + side effects from other tests
- Docs the tripleo folks pointed to:
- Tripleo quickstart logs:
  - https://gist.github.com/dianaclarke/2ebbaeb7e0454a5efc501207dd55a33f
- Review - nova: move tripleo-multinode to check -nv
- https://etherpad.openstack.org/p/nova-ocata-retrospective
- Red Hat Bug #1421633 - Any workaround of Bug #1444841 “Resize instance fails after creating host aggregate”
  - Resize instance fails after creating host aggregate
- Tripleo quickstart failed again after 132 minutes...
  - https://gist.github.com/dianaclarke/57a26f36d69111a2f1eef3c57ebd9ce4
- TODO: Set up the tripleo GUI
Week of February 6, 2017
- Red Hat Bug #1418614 - got an unexpected keyword argument 'prompt'
- How do you run the functional mysql db tests locally?
  - See MySQLOpportunisticTestCase, so: http://paste.openstack.org/show/603684/
  - Normally, these nova.tests.unit.db.test_migrations.TestNovaMigrationsMySQL test are skipped
  - After, they run and take about 24 second locally
  - Initial profiling dump: http://paste.openstack.org/show/597871/
- Profile migrations, consider changes like the following: http://paste.openstack.org/show/597627/
  - Before: there are 1122 COMMIT statements ($ grep COMMIT sql.out.fresh_db | wc -l)
  - Using the with migrate_engine.begin() idiom in migration 216, drops that number to 674
- Patch - Reduce the number of COMMIT & SELECT 1 statements
- How long does each migration take?
  - http://paste.openstack.org/show/598001/
- Which migrations are the slowest?
  - slowest: 216, 267, 232, 315, 228, 271, 245, 318, 314
  - http://paste.openstack.org/show/598003/
- And what about the api database migrations... How long to they take?
  - http://paste.openstack.org/show/598005/
- And data migrations? Where are those?
- Good to know - https://github.com/openstack/nova/blob/master/tools/db/schema_diff.py
- Patch - Fix nova-manage cell_v2 metavar strings
- Tripleo quickstart again...
  - this time on fedora 25 with a custom storage location
  - wow, 75m54.095s
- Create a stable/ocata devstack snapshot
- devstack nova client warnings? http://paste.openstack.org/show/598407/
- Red Hat Bug #1393243 - missing flavors from RHOS10 overcloud deployment
- Database dump: https://bugs.launchpad.net/nova/+bug/1622545/comments/6
- Red Hat #1384684 - [RFE][nova] Support flavor with no local disks
Week of January 30, 2017
- Changes based on review feedback - nova-manage docs for cell_v2
- Red Hat Bug #1416843 - metadata_workers controls number os nova_api processes
- Downstream Review - OSP9 - scheduler: Don't modify RequestSpec.numa_topology
- Downstream Review - OSP10 - scheduler: Don't modify RequestSpec.numa_topology
- Read - https://specs.openstack.org/openstack/tripleo-specs/specs/ocata/tripleo-composable-upgrades.html
- Watch - Add nova service support for composable upgrades
- https://etherpad.openstack.org/p/nova-ocata-rc1-todos
- Patch - Fix map_cell_and_hosts help
- Good to know - http://status.openstack.org/openstack-health/#/
- Review the Polarion test plans for simple tenant usage
- Review - doc: refer to the cell_v2 man pages from the cells v2 doc
- Review - doc: add some detail to the map_cell0 man page
- TODO: Is the right version of python-novaclient in brew for OSP11?
- TODO: Revisit the tripleo quickstart since that VM is b0rked again ;(
- Patch - Add offset & limit docs & tests
- Shadow the tripleo/nova:db:sync issues
- Shadow the tripleo issues in general:
  - pass: http://logs.openstack.org/15/424915/6/check/gate-puppet-openstack-integration-4-scenario001-tempest-centos-7/2c42212/
  - fail: http://logs.openstack.org/15/424915/8/check/gate-puppet-openstack-integration-4-scenario001-tempest-centos-7/b29f35b/
  - why?
- Red Hat Bug #1418614 - got an unexpected keyword argument 'prompt'
  - prompt added in: https://review.openstack.org/#/c/338614/ (Jul 2016, which is 2.10.0 min I think)
  - https://github.com/openstack/nova/blob/stable/mitaka/requirements.txt#L30 keystoneauth1>=2.1.0
  - https://github.com/openstack/requirements/blob/stable/mitaka/global-requirements.txt keystoneauth1>=2.1.0
  - https://github.com/openstack/requirements/blob/stable/mitaka/upper-constraints.txt keystoneauth1===2.4.2
- TODO: Red Hat Bug #1285359 - nova-manage db archive_deleted_rows doesn't work
- TODO: Lee's downstream reviews
- What's InfraRed? https://media.readthedocs.org/pdf/infrared/latest/infrared.pdf
- Profile migrations, consider changes like the following: http://paste.openstack.org/show/597627/
- How do you run the mysql db migration tests locally?
Week of January 23, 2017
- Bug #1658148 - Horizon/Nova interactions broken
  - See: Pass relevant parameters to Token based authentication
  - See: [CLI] Fix token auth type
- Watch - Implement a proper cell_v2 setup
- Circle back to this: https://github.com/openstack/tripleo-quickstart - [notes]
- Bug? Quickstart docs say >= Fedora 22 is fine. But perhaps they expect you to run it without the distro check for Fedora... not sure.
  - https://github.com/openstack/tripleo-quickstart/commit/08a98c00e772426fc157b768d5b521e3137fce0d
- Downstream Review - Refresh stale volume BDMs in terminate_connection
- Watch the upgrades part of the tripleo bootcamp videos: https://mojo.redhat.com/docs/DOC-1122111
- Audit the cells v2 man page patches for changes that since last week
  - I think this is the only difference: https://review.openstack.org/#/c/420132/4/nova/cmd/manage.py
- Minor edits to patch series - nova-manage docs for cell_v2
- Review - Try to use 2.40 microversion for simple-tenant-usages
- Review - Fix doc generation warnings
- Red Hat Bug #1416843 - metadata_workers controls number os nova_api processes
  - conf for metadata_workers: https://github.com/openstack/nova/blob/f9d7b383a7cb12b6cd3e6117daf69b08620bf40f/nova/conf/service.py#L144
  - etc/nova/nova.conf.sample:#metadata_workers = <None>
  - conf for osapi_compute_workers: https://github.com/openstack/nova/blob/f9d7b383a7cb12b6cd3e6117daf69b08620bf40f/nova/conf/service.py#L112
  - etc/nova/nova.conf.sample:#osapi_compute_workers = <None>
  - http://codesearch.openstack.org/?q=osapi_compute_workers&i=nope&files=&repos=
  - http://codesearch.openstack.org/?q=metadata_workers&i=nope&files=&repos=
  - focus on: openstack / tripleo-heat-templates, openstack / puppet-nova, and openstack / instack-undercloud
  - Tune number of workers for undercloud - https://review.openstack.org/#/c/386696/
  - oh! https://github.com/openstack/nova/blob/f9d7b383a7cb12b6cd3e6117daf69b08620bf40f/nova/service.py#L319
  - what my devstack looks like wrt workers and nova-api processes: http://paste.openstack.org/show/596645/
  - a nova-api service is started for each enabled_apis (['osapi_compute', 'metadata']): https://github.com/openstack/nova/blob/d7b5be1e3c1c0e5c6ae429173967aea55c2f0609/nova/cmd/api.py#L56
- Downstream Review - OSP9 - scheduler: Don't modify RequestSpec.numa_topology
- Downstream Review - OSP10 - scheduler: Don't modify RequestSpec.numa_topology
- TODO: Review - Reduce time and memory usage of tenant usage.
- TODO: Now what? http://docs.openstack.org/developer/tripleo-docs/
- TODO: Is there anything from my tripleo quickstart notes that's actionable? That is, fix the failing dependencies etc in the Fedora case?
Week of January 16, 2017
- Start here: https://github.com/openstack/tripleo-quickstart
- Red Hat Bug #1403728 - Instance not resizing, skipping migration
- Read: https://review.openstack.org/#/c/420198/11/doc/source/cells.rst
- Read: https://wiki.openstack.org/wiki/Nova-Cells-v2
- Review - Add nova-manage cell_v2 list_cells command
- Review - Add nova-manage cell_v2 delete_cell command
- Review - Add --verbose option to discover_hosts command
- Review - Fix and add some notes to the cells v2 first time setup doc
- Review - Create HostMappingList object
- Review - Add InstanceMappingList.get_by_cell_id
- Shadow - Red Hat Bug #1405098 - Ocata requires additional cellvs2 setup
- Shadow - Red Hat Bug #1406908 - Compute Placement API Upgrade Considerations
- Patch - nova-manage docs: add cells commands prep
- Patch - nova-manage docs: cell_v2 map_instances
- Patch - nova-manage docs: cell_v2 map_cell_and_hosts
- Patch - nova-manage docs: cell_v2 verify_instance
- Patch - nova-manage docs: cell_v2 simple_cell_setup
- Patch - nova-manage docs: cell_v2 map_cell0
- Patch - nova-manage docs: cell_v2 discover_hosts
- Patch - nova-manage docs: cell_v2 create_cell
- Patch - nova-manage docs: cell_v2 list_cells
- Patch - nova-manage docs: cell_v2 delete_cell
- TODO: What's the correct ordering of the cells commands for a fresh install vs an upgrade?
- Debug - update constraint for python-novaclient to new release 7.0.0
- Patch - Fake deprecated nova client properties
- Patch - Fix map_cell_and_hosts help
- How does devstack uses the cells v2 commands?
  - https://github.com/openstack-dev/devstack/blob/eb9d42b65c8dbbe531e8124b442f13db5dff2993/lib/nova#L948
  - https://github.com/openstack-dev/devstack/search?utf8=%E2%9C%93&q=cell_v2
  - Right now it's using: map_cell0, simple_cell_setup, and map_cell_and_hosts
- How does grenade uses the cells v2 commands?
  - https://github.com/openstack-dev/grenade/search?utf8=%E2%9C%93&q=cell_v2
  - Right now it's using: map_cell0, simple_cell_setup, and verify_instance
- How does tripleo uses the cells v2 commands?
  - http://codesearch.openstack.org/?q=%2Fusr%2Fbin%2Fnova-manage&i=nope&files=&repos=
  - So far I just see: simple_cell_setup
- Red Hat Bug #1412197 - Nova misconfigured after upgrade to RHOS-11: ConfigFileValueError: Value for option osapi_compute_workers is not valid
  - Tip: see Red Hat Bug #1412203 - Nova misconfigured after upgrade to RHOS-11: ConfigFileValueError: Value for option metadata_workers is not valid
  - Bug #1633735 - Can't deploy overcloud with ocata/trunk on rhel
  - This error is mentioned here: http://eavesdrop.openstack.org/irclogs/%23tripleo/%23tripleo.2016-11-30.log.html#t2016-11-30T15:23:29
  - Bug #1631133 - nova's osapi_compute_workers is not valid (must be greater than 1)
  - http://codesearch.openstack.org/?q=osapi_compute_workers&i=nope&files=&repos=
  - Introduced here: https://review.openstack.org/#/c/332861/13/nova/conf/service.py
  - Hmmm... maybe NovaWorkers should default to one, not zero: http://codesearch.openstack.org/?q=NovaWorkers&i=nope&files=&repos=
  - Only set NovaWorkers in the non-default case, merged Oct 2016
  - Add apache workers to nova-api conditional, merged Oct 2016
  - Perhaps the heat template needs an upgrade_tasks section that removes the nova::api::osapi_compute_workers config if it is set to zero.
  - see this upgrades WIP: https://review.openstack.org/#/c/405241/10/puppet/services/nova-api.yaml
- Bug #1658148 - Horizon/Nova interactions broken
Week of January 9, 2017
- Help with reviews here: https://blueprints.launchpad.net/nova/ocata
- Review - Simple tenant usage pagination - Horizon
- Bug #1654806 - Multiple possible networks found, use a Network ID to be more specific
- OpenStack Terms
- Start again with first principles: http://docs.openstack.org/admin-guide/
- Patch - SQL-Alchemy -> SQLAlchemy
- Quick lookup: http://docs.openstack.org/admin-guide/common/glossary.html
- Patch - WIP - Add nova microversion 2.40 support
- Current nova code coverage: 88%
- Red Hat Bug #1403728 - Instance not resizing, skipping migration
  - http://www.danplanet.com/blog/2016/03/03/evacuate-in-nova-one-command-to-confuse-us-all/
- Bug #1655710 - Query parameter validation using json schema - error messages
- Patch - docs - Connect to placement service & retries
- TODO: read all of these: http://www.danplanet.com/blog/tag/nova/
- Patch - Query parameter validation for usages API
- Patch - Query parameter validation for hypervisors API
- Watch the TripleO bootcamp videos: https://mojo.redhat.com/docs/DOC-1122111
  - In particular: https://docs.google.com/presentation/d/1HG4vgTGV5AwvSnZNOHTCTABSshR2sZHoN2FEgarOT78/edit#slide=id.ge7e47c999_4_58
- Shadow - Red Hat Bug #1405098 - Ocata requires additional cellvs2 setup
- Shadow - Red Hat Bug #1406908 - Compute Placement API Upgrade Considerations
- Patch - copy pasta error
- new non-devstack experimental CI job
  - gate-tripleo-ci-centos-7-nonha-multinode
  - http://lists.openstack.org/pipermail/openstack-dev/2016-October/105080.html
Week of January 2, 2017
- Holiday Monday
- Patch - Missing usage next links in api-ref
- Patch - api-ref: use the examples with paging links
- More manual testing of the STU changes against devstack
  - http://paste.openstack.org/show/593827/
  - Unable to determine DISTRO, can not continue.
  - problems with: dnf install redhat-lsb-core
- Review - Add CLI to show instance usage audit logs
- Review - Add batch add/remove server tags support
- Review - Add evacuate host method
- Review - tests: Don't reinvent init
- Review - tools: Remove 'colorizer'
- Review - tools: Remove 'with_venv'
- Review - tools: Remove 'install_venv', 'install_venv_common'
- Review - tools: Remove 'clean-vlans'
- Review - tools: Remove 'enable-pre-commit-hook'
- Memorize this, stupid brain: tox -e pep8 -- -HEAD
- Red Hat Bug #1406356 - live_snapshot can be enabled when RBD snapshot/clone is performed
  - stable/liberty (not merged): https://review.openstack.org/#/c/287954/
  - master (mitaka, merged): https://review.openstack.org/#/c/205282/
  - Downstream only - rbd: use rbd snapshots and cloning for ephemeral disk snapshots
- Review - Use JSON-Schema to validate query parameters for keypairs API
- Patch - Query parameter validation for usages API
- Patch - Query parameter validation for hypervisors API
- Good trick to remember (field): nova list --field host,status,name
Week of December 26, 2016
- Slow week, vacation
- Red Hat Bug #1408707 - Instances using SRIOV ports are not getting spawned.
- Patch - api-ref: microversion 2.40 overview
- Patch - Blacklist rather than whitelist autodoc modules
- What's up with all the sphinx warnings in the client docs? New? Anyhoo... fix them.
  - Turns out this was just old/bad files from previous doc builds
- Bug #1533087 - there is useless 'u' in the wrong info when execute a wrong nova command
  - Dig into the root cause before +1'ing the replace(" 'u", " '") solution
  - Review - Extra 'u' in the error message for nova commands
- Review - Remove HTTP error code when creating flavor
- Patch - Remove Rules.load_json warning
Week of December 19, 2016
- Slow week, vacation
- Client Patch - Microversion 2.40 - Simple tenant usage pagination
- Patch - osapi_max_limit -> max_limit
- Review - Add some missing modules in API reference
- Review - Remove v2.40 from URL string in usage API docs
- Patch - Missing usage next links in api-ref
Week of December 12, 2016
- Client Patch - Microversion 2.40 - Simple tenant usage pagination
- Patch - Simple tenant usage pagination
- Manual testing of STU
  - try again later in the week; there are too many unrelated errors with cells & transport_url, etc
  - even after rebasing on the nova change, I'm getting "oslo_config.cfg.NoSuchOptError: no such option transport_url in group" in devstack
  - Undercloud upgrade fails with "Cell mappings are not created, but required for Ocata"
- Review - microversion 2.39 deprecates image-metadata proxy API
- Review - Replaces uuid.uuid4 with uuidutils.generate_uuid()
- Patch - Do not attempt volume swap when guest is stopped
- Review - Horizon: Simple tenant usage pagination
- Release notes job failures - Use liberty-eol tag for liberty release notes
- Patch - Add offset & limit docs & tests
- Patch - osapi_max_limit -> max_limit
- Review - Port test_simple_tenant_usage to Python 3
- Review - Port security group related tests to Python 3
- Review - Add soft delete wrinkle to api-ref
- Patch - Failing test (mac osx) - test_cache_ephemeral
- Red Hat Bug #1403728 - Unexpected "Instance not resizing, skipping migration" caused by nova evacuation
- Review - Fixed the ne implementation in base.Resource
- Review - Microversion 2.40 - Enable cold migration woth target host
Week of December 5, 2016
- Revert the revert ;( - Microversion 2.39 - Simple tenant usage pagination
- Functional tests for STU in python-novaclient
- Trying to figure out why python-novaclient gate is failing
  - Bug #1647464 - test_list_all_servers broken since at least 12/2
- TODO: Is --debug working properly for nova usage?
- Review - Fix doc generation errors
- Review - Fix can't process the resource with name 'help'
- Review - Handle MarkerNotFound from cell0 database
- Review - Removing version error from command nova keypair-add help
- Patch - CONF.osapi_max_limit -> CONF.api.max_limit
- Patch - Usage missing from generated docs
- Review - Introduce helper for checking args deprecation
- Review - Use more specific asserts in tests
- Review - Remove redundant args of _construct_http_client
- Review - Sort arguments for client's methods
- Review - Restict usage *args for novaclient.client.Client
- Microversion 2.39 is taken ;( Bump to 2.40
- Review - remove variable '_' from tests
- Review - Add min-disk and min-ram filters to list flavors
- Review - Make SessionClient interface similar to HTTPClient
- Review - Deprecate conection_pool variable
- Run coverage on the usage endpoint & increase unit test coverage
- Review - Fix crashing during guest config with pci_devices=None
- Review - Add paging support in server groups list
- Fix the python 3 issues with the existing tests (they are in tests-py3.txt)
Week of November 28, 2016
- Client Patch - Microversion 2.39 - Simple tenant usage pagination
- Patch - Simple tenant usage pagination
- TODO: Add tests for the endpoints that use limit & offset
- Patch - Correct copy/paste errors in help
- Backport - Correct copy/paste errors in help
- simple tenant usage notes
- Hours fighting with devstack
- Red Hat Bug #1399096 - after editing a flavor used by an instance, flavor is not found when doing nova show
  - known issue, able to repo, seen this reported a few times now, pretty certain it's wontfix
  - https://review.openstack.org/#/q/topic:bp/flavor-cell-api
  - see mailing list: "Should flavors in the API DB for cells v2 be soft-deletable?"
  - https://github.com/dianaclarke/openstack-notes/wiki/wrong-flavor
- Trying to fix these errors: http://logs.openstack.org/79/319379/36/check/gate-nova-python27-db-ubuntu-xenial/9a17286/console.html
Week of November 21, 2016
- Compliance and Ethics Training/Certification
- Red hat Bug #1299980 - Pagination and Filtering For Listing OpenStack Server Groups Using REST API
  - server group notes
- Client Patch - Microversion 2.39 - Simple tenant usage pagination
- Patch - api-ref: add missing os-server-groups parameters
- TODO: Audit api-ref for missing limit & offset query parameter documentation
- Read: https://blueprints.launchpad.net/nova/+spec/flavor-root-disk-none (and spec etc)
- Red Hat Bug #1393124 - openstack-nova-conductor-12.0.3-1.el7ost.noarch requires python2-cryptography-1.3.1-3.el7.x86_64
  - Issues on upgrading to RHEL 7.3 with OSP8/9
- Patch - Simple tenant usage pagination
- Red Hat Bug #1352922 - [RFE] Simple Tenant Usage Pagination Support
Week of November 14, 2016
- Spec - Simple tenant usage pagination
- Update "Preventative Action" for the paramiko bug
- TODO: Simple tenant usage sorting
- TODO: Functional tests for STU in python-novaclient
- TODO: Update api-ref & add reno notes (done for python-novaclient)
- Client Patch - Microversion 2.39 - Simple tenant usage pagination
- Red Hat Bug #1393124 - openstack-nova-conductor-12.0.3-1.el7ost.noarch requires python2-cryptography-1.3.1-3.el7.x86_64
  - Waiting on needinfo
- TODO: Do we need a new database index for simple tenant pagination?
- TODO: Manually test the nova & python-novaclient changes against devstack
- TODO: Hmmm... should I be doing something for python-openstackclient too (for simple tenant usage paging)?
  - I think so: https://github.com/openstack/python-openstackclient/blob/master/openstackclient/compute/v2/usage.py
- Red Hat Bug #1385084 - 404 in nova-api Unexpected exception in API method
  - waiting on networking DFG, pinged
- Red Hat Bug #1387138 - nova allows user to boot an instance while specifying a non-existing availability zone
  - waiting on needinfo, pinged
- Patch - Bump client microversion to 2.38
- Review - Consistent query parameters validation
- Review - Add whitelist for filter and sort query parameters for server list API
- http://stackoverflow.com/questions/5631078/sqlalchemy-print-the-actual-query
- Bug #1640646 - Tempest runs volume snapshot tests even when volume snapshots are disabled
- Patch - Database poison note
Week of November 7, 2016
- Spec - Simple tenant usage pagination
- Patch - Simple tenant usage pagination
- Client Patch - Microversion 2.39 - Simple tenant usage pagination
- Review - Fix 1 of 2 database poison warnings in resource providers
- TODO: Preventative action thing
- Review - Fix 1 of 2 database poison warnings in resource providers
- What was microversion 2.38? python-novaclient impact? Don't think so.
  - Return HTTP 400 on list for invalid status
- How do you add reno notes? Answer: nova docs
- Patch - Bump client microversion to 2.38
- Patch - Change database poison warning to an exception
- Patch - Fix database poison warnings, part 25
- Bug #1640646 - Tempest runs volume snapshot tests even when volume snapshots are disabled
- Red Hat Bug #1393124 - openstack-nova-conductor-12.0.3-1.el7ost.noarch requires python2-cryptography-1.3.1-3.el7.x86_64
- http://docs.openstack.org/developer/devstack/configuration.html#service-repos
Week of October 31, 2016
- slow week, sick
- Patch - tests for https://review.openstack.org/#/c/382449/
- Spec - Simple tenant usage pagination
- Patch - Simple tenant usage pagination
- TODO: update client for usage, example
- Red hat Bug #1390054 - nova.api.openstack.extensions Unexpected exception in API method The resource could not be found
- Patch - EventReporterStub
- Last poison warnings: http://paste.openstack.org/show/587184/
Week of October 24, 2016
- Devstack (stable/newton) - Unknown database 'nova_api': http://paste.openstack.org/show/586710/
  - Looks like this was reported already: https://bugs.launchpad.net/devstack/+bug/1620028
  - Does this fix it? https://review.openstack.org/#/c/384916/ I think so.
- Downstream bug triage - compute backlog
- Red Hat Bug #1341401 - nova-api memory usage incredibly high
- Bug #1635467 - _emit_versioned_exception_notification exception
  - Start here: Transform wrap_exception notification to versioned format
  - Patch - Fix exception raised in exception wrapper
- Patch - Simple tenant usage pagination
- https://etherpad.openstack.org/p/ocata-summit-xp-scaling-review-teams
- https://wiki.openstack.org/wiki/Design_Summit/Ocata/Etherpads#Nova
- Red Hat Bug #1373473 - Migration for instance refers to another host's instance!
- Red Hat Bug #1388417 - Failed to detach an encrypted volume
  - no luck adding any value here
- Patch - Fix SAWarning in TestResourceProvider
- Current database poison warnings: http://paste.openstack.org/show/587194/ & http://paste.openstack.org/show/587184/
- Red Hat Bug #1389294 - nova resize does not resize the storage for instance booted from an image
- simple tenant usage notes
- Patch - Fix database poison warnings, part 23
- Patch - Fix database poison warnings, part 24
- TODO: Review - Reduce time and memory usage of tenant usage
Week of October 17, 2016
- Spec - Simple tenant usage pagination
- Review some mox patches
- Patch: Remove paramiko dependency
- Red Hat Bug #1373473 - Migration for instance refers to another host's instance!
  - Downstream Patch: Change log level to debug for migrations pairing - OSP9
- WIP: Add a nova unit test that asserts that SSH key ASN.1 encoding is DER (not BER)
- Red Hat Bug #1373318 - Can't re-start a stopped instance after trying to volume-update on it
  - Start here: Save bdm in swap_volume
  - stable/newton: http://paste.openstack.org/show/586579/ & http://paste.openstack.org/show/586592/
  - stable/mitaka: http://paste.openstack.org/show/586586/ & http://paste.openstack.org/show/586591/
  - If you check that the guest is active first: http://paste.openstack.org/show/586630/
  - Patch - Do not attempt volume swap when guest is stopped
- Red Hat Bug #1387138 - nova allows user to boot an instance with specifying a inexistence availability zone
  - Start here: Parse availability_zone in API
  - Patch - Always validate availability_zone on 'nova boot'
- Downstream bug triage - compute backlog
- Patch - EventReporterStub
Week of October 11, 2016
- https://blueprints.launchpad.net/nova/+spec/paginate-simple-tenant-usage
- simple tenant usage notes
- See: "API changes on limit / marker / sort in Newton" thread on openstack-dev
- TODO: python-novaclient needs to be updated too (for simple tenant usage paging)
- Patch - EventReporterStub
- Bug #1632538 - Using generate_service_certificate and undercloud_public_vip in undercloud.conf breaks nova
- Patch - Simple tenant usage pagination
- Red Hat Bug #1324778 - 404 in nova-api Unexpected exception in API method
- Spec - Simple tenant usage pagination
Week of October 3, 2016
- https://blueprints.launchpad.net/nova/+spec/paginate-simple-tenant-usage
- simple tenant usage notes
- Bug? nah... looks like it's fixed already
  - stable/mitaka devstack: WHERE migrations.deleted = 0 AND migrations.status IN ('e', 'r', 'r', 'o', 'r')
  - stable/newton devstack: WHERE migrations.deleted = 0 AND migrations.status IN ('error')
- Make these tests pass on a mac osx again: traceback 1 & traceback 2
  - Patch - Patch mkisofs calls
  - Patch - ENOENT error on '/dev/log'
- http://codesearch.openstack.org/
Week of September 26, 2016
- Patch - Fix database poison warnings, part 15
- Patch - Fix database poison warnings, part 16
- Patch - Fix database poison warnings, part 17
- Patch - Fix database poison warnings, part 18
- Patch - Fix database poison warnings, part 19
- Patch - Fix database poison warnings, part 20
- Patch - Fix database poison warnings, part 21
- Patch - Fix database poison warnings, part 22
- Patch - EventReporterStub
- Reviewed: Don't load deleted instances
- Reviewed: Remove mox from unit/api/openstack/compute/test_migrations.py
- More setting up new machine... VPN etc
- Creating devstack snapshots - notes
- RDO Test Day - notes
- https://blueprints.launchpad.net/nova/+spec/paginate-simple-tenant-usage
- Current test failures: http://paste.openstack.org/show/583629/
- 100 slowest tests: http://paste.openstack.org/show/583636/
  - main culprit: api: use sane default in wrap_db_retry()
- Patch: oslo_db wrap_db_retry defaults changed
- Reviewed: Prevent us from sleeping during DB retry tests
Week of September 19, 2016
- Sick, slow week.
- Red Hat Bug #1373613 - Backport Nova "AIO mode for disk devices" patches to Liberty (RHOS-8) from Mitaka (RHOS-9)
  - aio notes
- Have the mac os x failures in the unit tests gone away? Checking... nope
  - TODO: circle back to this
- Watched: Retrospective in 5 Minutes by Elise Gafford
- Reviewed: stub out instances_path in unit tests
- Reviewed: Use cursive for signature verification
- Setting up new machine - notes
- Patch - Fix database poison warnings, part 14
- Red Hat Bug #1373473 - Migration for instance refers to another host's instance!
  - tips: "should distinguish between case of the instance being deleted and being from the wrong host", "suppress log message if instance known to be deleted or mark the migration record as deleted"
- What should I start reviewing upstream? Perhaps:
  - https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/remove-mox-ocata
  - https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/goal-python35
- Red Hat Bug #1378008 - Moving haproxy galera proxy VIP result in nova controller service to be down until sql timeout reached
- zzzeek explains database settings & openstack
  - See also on openstack-dev "Excessively high greenlet default + excessively low connection pool defaults leads to connection pool latency, timeout errors, idle database connections / workers"
- TODO - Read: haalchemy
- Reviewed: Remove mox from api/openstack/compute/test_availability_zone.py
- Reviewed: Stop using mox from compute/test_user_data.py
- Reviewed: Stop using mox from unit/api/openstack/test_wsgi.py
- Reviewed: Remove mox from compute/test_server_metadata.py
Week of September 12, 2016
- Shadowing upstream release candidate issues
- Red Hat Bug #1373554 - ssh-keygen-to-Paramiko change breaks third-party tools
  - ASN.1 DER vs BER notes
  - Downstream Patch: Remove paramiko dependency - OSP8
- Red Hat Bug #1376493 - ssh-keygen-to-Paramiko change breaks third-party tools
- Downstream Patch: Remove paramiko dependency - OSP9
- Bug #1613199 - nova should accept pre-pended comments in public keys
- Note that the prepended comment doesn't change the fingerprint - ssh public key comments notes
- Red Hat Bug # 1375048 - Volume's info gets inconsistent between nova and cinder, when two instances are booted from the same volume simultaneously
  - Detach and terminate conn if Cinder attach fails
  - Add checks for preventing unexpected detaching volume
- Red Hat Bug #1373613 - Backport Nova "AIO mode for disk devices" patches to Liberty (RHOS-8) from Mitaka (RHOS-9)
  - libvirt: use native AIO mode for image backends
    - landed upstream master circa January 2016 (Mitaka)
    - Downstream Patch: libvirt: use native AIO mode for image backends
  - libvirt: use native AIO mode for cinder volumes
    - landed upstream master circa January 2016 (Mitaka)
    - Downstream Patch: libvirt: use native AIO mode for cinder volumes
Week of September 6, 2016
- Installed Fedora 24 on new work machine
- Downstream bug triage - compute backlog
- TODO: https://www.rdoproject.org/testday/newton/milestone3/
  - RDO test days: September 8th and 9th, 2016
  - Skipped: Ended up working on Red Hat Bug #1373554 instead.
- Red Hat Bug #1373318 - Can't re-start a stopped instance after trying to volume-update on it
  - volume update notes
  - OUTCOME: I was only able to partially reproduce this one, so I left the status as-is for another SME.
- Red Hat Bug #1373554 - ssh-keygen-to-Paramiko change breaks third-party tools
  - Bug #1483132 - ssh-keygen-to-Paramiko change breaks third-party tools
  - Bump paramiko to version 2.0
  - OUTCOME: Downstream backport patches up for review.
- Patch: Remove paramiko dependency
- Downstream Patch: Remove paramiko dependency - OSP9
- Downstream Patch: Remove paramiko dependency - OSP8
- Triage: Red Hat Bug #1373613 - Backport Nova "AIO mode for disk devices" patches to Liberty (RHOS-8) from Mitaka (RHOS-9)
  - libvirt: use native AIO mode for image backends
  - libvirt: use native AIO mode for cinder volumes
Week of August 29, 2016
- Switch focus to the downstream compute bugzilla backlog
- Red Hat Bug #1368588 - Permission denied: '/var/log/nova/nova-manage.log
  - OUTCOME: DUPLICATE, Marked as a duplicate of Red Hat Bug #1289761 - PackStack installs Nova crontab that nova user can't run
- Red Hat Bug #1350424 - compute_monitors parameter is a nested list
  - OUTCOME: NOTABUG, Closed since the need info flag was 2 months old, and all interested parties seem to have moved on.
- Red Hat Bug #1340151 - InstanceInfoCacheNotFound: Info cache for instance not found error in nova-compute logs
  - OUTCOME: NOTABUG, Closing since the original reporter was unable to reproduce the issue and volunteered to reopen the bug in the future if the issue arises again.
- Red Hat Bug #1371624 - Nova flavor 'Is_Public' remains 'True' in 'nova flavor-list' output after tenant access applied
  - OUTCOME: NOTABUG, Closed since it looks like the workflow was just incorrect.
- PyMySQL 0.7.7 is injecting strange codes into exception messages
- Red Hat Bug #1372290 - starting nova-api timesout using certain sudoers configuration
  - OUTCOME: NOTABUG, Closed since it appears to be more of a troubleshooting notepad.
- Unboxed new work machine, started setting it up.
Week of August 22, 2016
- LinuxCon & KVM Forum
- Compute DFG - Trello
- (╯°□°）╯︵ ┻━┻
- TODO: Add more documentation around:
  - Inconsistency quota size nova, neutron, openstack command
- TODO: Finish watching the GSS training videos
- TODO: Matt's new tests don't pass on Mac OS X (and they should according to the OpenStack docs)
Week of August 15, 2016
- Vacation
Week of August 8, 2016
- Vacation
Week of August 1, 2016
- Vacation
Week of July 25, 2016
- Subteam reviews
- Patch - Fix database poison warnings, part 13
- Patch - Fix database poison warnings, part 14
- Patch - Change database poison warning to an exception
- 51 database poison warnings left in master - tests with warnings
- Rabbit hole: What are valid config_drive values?
- Patch babysitting while Matt is on vacation:
  - https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:libvirt-imagebackend
- Red Hat Bug #1358099 - novncproxy: directory listing on :6080/include/
  - That's the directory listing for: kanaka/noVNC/include/ (noVNC: HTML5 VNC Client)
  - Nova Bug #1116184 - nova-novncproxy allows directory listing
    - closed invalid, circa 2013
  - noVNC Bug - nova-novncproxy allows directory listing
    - closed fixed, circa Sept 2014
  - Nova Bug #1447675 - directory listing of the service No-VNC
    - fix released, OpenStack Compute (nova) 12.0.0 "liberty", circa Oct 2015
  - Mirantis OpenStack Bug - Block directory listing for noVNC service
    - fixed "Mirantis OpenStack 7.0-mu-2", circa Feb 2016
  - Relevant version bump - Bump websockify to 0.6.1
    - global requirements updated Aug 2015
  - Websockify:
    - Changelog: fix in version 0.6.1, circa May 2015
    - https://github.com/kanaka/websockify/commits/v0.6.1
    - Patch - Restore functionaltiy to file_only.
Week of July 18, 2016
- Missing local root disk (see: http://lists.openstack.org/pipermail/openstack-dev/2016-June/097529.html) - [notes]
- Verify the coverage job is working for nova, mailing list thread
- Patch - Add nova-coverage-db to check queue
- Following along from home: https://etherpad.openstack.org/p/nova-newton-midcycle
- Reading up on rdopkg (re: questions about OSP6)
- Subteam reviews
- Patch - Fix database poison warnings, part 9
- Patch - Fix database poison warnings, part 10
- Patch - Fix database poison warnings, part 11
- Patch - Fix database poison warnings, part 12
- Looks like these are the last of the poison warnings: http://paste.openstack.org/show/539315/
  - nova.tests.unit.virt.xenapi.test_xenapi
  - nova.tests.unit.virt.vmwareapi.test_driver_api
- Overview of the libvirt instance storage series
  - http://lists.openstack.org/pipermail/openstack-dev/2016-July/099858.html
Week of July 11, 2016
- Slow week, sick
- Subteam reviews
- Shadowing gate failure debugging
- Watched the video from the first team retro
- First tempest attempt - verify root/swap/ephemeral disk sizes
- Missing local root disk (see: http://lists.openstack.org/pipermail/openstack-dev/2016-June/097529.html) - [notes]
- Question - Why is the local root disk missing ('/dev/vda')?
- Question - How many patches are abandoned for each OpenStack project? Start here:
  - http://russellbryant.net/openstack-stats/
  - http://git.openstack.org/cgit/openstack-infra/reviewstats/
  - https://github.com/openstack-infra/reviewstats
    - $ export GERRIT_USER=diana-clarke
    - $ ./genresults-openreviews.sh
    - $ cd results/
  - https://review.openstack.org/Documentation/user-search.html
Week of July 4, 2016
- Subteam reviews
- Order testing machine
- First tempest attempt - verify root/swap/ephemeral disk sizes
- Increase tempest coverage for image backends - [notes]
- Tempest ssh timeouts:
  - Local example: http://paste.openstack.org/show/527031/
  - CI example: citrix-xenserver [full log]
  - SSH Client hangs on execute command
  - SSHTimeout in tempest trying to verify that computes are actually functioning
- Trying to help with the gate-tempest-dsvm-postgres-full errors
  - example pass: http://logs.openstack.org/89/339289/2/check/gate-tempest-dsvm-postgres-full/cec9383/console.html.gz
  - example fail: http://logs.openstack.org/89/339289/1/check/gate-tempest-dsvm-postgres-full/5c9aa4a/console.html.gz
Week of June 27, 2016
- Subteam reviews
- Order testing machine
- Increase tempest coverage for image backends - [notes]
- Resizing disks - [notes]
- Resize Ephemeral Disk Bug - https://notehub.org/o3mjz
- Started watching the GSS training videos
Week of June 20, 2016
- Subteam reviews, answering reviews, rebasing this patch series, etc
- Red Hat Bug #1346236 - Error unable to retrieve hypervisor information is coming in horizon dashboard
- Re-visit resize of existing disks (not via the create_ paths)
  - Patch: https://review.openstack.org/#/c/332416/
- Review CI coverage, in particular: flat & ploop & lvm
  - Are these flat CI errors related?
    - I don't think so
  - Virtuozzo CI errors:
    - n-cpu logs
    - traceback: http://paste.openstack.org/show/521172/
      - No such file or directory: ploop grow -s 3145728K /opt/stack/data/nova/instances/46555273-589a-460d-9117-def64261ba65/disk/DiskDescriptor.xml
    - fixed, CI caught legit error in my ploop impl
  - LVM: lvm job in the experimental queue is running fine
  - experimental queue errors: n-cpu logs
    - TODO
- Re-visit the original spec as a refresher
- Increase tempest coverage for image backends - [notes]
- TODO: Review Matt's recent patch series
- Current unit test times: http://paste.openstack.org/raw/520831/
- An OpenStack Compute (Nova) Deep Dive with John Garbutt (johnthetubaguy) - https://www.youtube.com/watch?v=zyJYC8ak_88
Week of June 13, 2016
- Subteam reviews, answering reviews, rebasing this patch series, etc
- Revisit missing backing files for the nth time
- These tests are using the old methods still: http://paste.openstack.org/show/515777/
- image backend checklist
- Read - http://lists.openstack.org/pipermail/openstack-dev/2016-June/097529.html
- Red Hat Bug #1346236 - Error unable to retrieve hypervisor information is coming in horizon dashboard
  - Juno code:
- Virtuozzo Errors:
  - June 8th (before is not check was added)
    - screen-g-api logs
      - GET /v1/images/ HTTP/1.1" 404
    - screen-n-cpu logs
      - ERROR nova.compute.manager Image could not be found.
  - June 10
    - screen-n-cpu logs
      - Invalid disk path: sudo nova-rootwrap /etc/nova/rootwrap.conf prl_disk_tool resize --size 3072M --resize_partition --hdd /opt/stack/data/nova/instances/b3991156-71e8-4794-ad5c-1b64d83b8af7/disk
      - full traceback: http://paste.openstack.org/show/516034/
      - http://lists.openstack.org/pipermail/openstack-dev/2016-June/097371.html
- Flat CI:
  - TODO DNM: use libvirt.images_type=flat by default
- TODO "mriedem: any idea on how hard it would be to add a functional test for https://review.openstack.org/#/c/279669/"
- TODO: https://github.com/openstack/tempest
Week of June 6, 2016
- image backend checklist
- Review subteam patches [start here]
- Downstream bug cleanup:
  - Red Hat Bug #1330628 - no retry when using 'nova image-list' on next configured glance api_servers endpoint
    - moved to MODIFIED (stable liberty release was cut upstream)
  - Red Hat Bug #1329018 - Request for Backport to use Port Group and Key in binding details in vmwareapi
    - answered QA
  - Red Hat Bug #1344173 - nova-compute failed to cleanup deleted instances
    - triage: requested access to parent bug (cloned from kilo bug)
    - now with access: Red Hat Bug #1337471 - nova-compute failed to cleanup deleted instances
    - perhaps related? Bug #1412285 - InstanceInfoCacheNotFound exception
Week of May 30, 2016
- image backend checklist
- High-level comparison of image backend create methods: http://paste.openstack.org/show/507360/
- Patch - Add check_backing_from_func & check_backing_from_image
- Patch - Delete old imagebackend methods & tests
  - known failing tests as a result (until rebase onto Matt's branch) http://paste.openstack.org/show/506351/
- Patch - Remove the last bits of mox from test_imagebackend
- Red Hat Bug # 1340996 - Block Device Mapping corrupt occasionally during volume attaching
  - For bug triage:
    - Race condition in quickly attaching / deleting volumes
    - Unnecessary bdm entry is created if same volume is attached twice to an instance
Week of May 23, 2016
- image backend checklist
- Patch - Qcow2: add create_from_image & create_from_func
- Patch - LVM: add create_from_image & create_from_func
- Started on check_backing_from_func & check_backing_from_image
- Started to delete the old imagebackend methods and audit the resulting test failures.
Week of May 16, 2016
- image backend checklist
- Patch - Ploop: add create_from_image & create_from_func
- Patch - Rbd: add create_from_image & create_from_func
- Patch - Flat: add create_from_image & create_from_func
- Patch - Image: add create_from_image & create_from_func
- http://docs.openstack.org/infra/manual/developers.html#adding-a-dependency
- Downstream review: Retain instance metadata for 'evacuate' on shared storage
- Red Hat Bug #1313254 - no retry when using 'nova image-list' on next configured glance api_servers endpoint
  - Pinged upstream about stable releases. Answer: stable releases (liberty & mitaka) will happen around n-1
- Red Hat Bug #1331286 - LibvirtError occurs during evacuation when nfs volume is used
  - Pinged mdbooth & danpb about the upstream review
- Red Hat Bug #1329018 - Request for Backport to use Port Group and Key in binding details in vmwareapi
  - packaged openstack-nova-12.0.3-4.el7ost (OSP8/Liberty)
- Red Hat Bug #1329250 - Request for Backport to use Port Group and Key in binding details in vmwareapi
  - packaged openstack-nova-13.0.0-5.el7ost (OSP9/Mitaka)
Week of May 9, 2016
- Review, OSP8/Liberty backport - Retain instance metadata for 'evacuate' on shared storage
- Briefly revisited the metadata cache & memcached issue - [notes]
  - TODO: update notes to recap discussions
- Storage pools - [notes]
- Red Hat Bug #1330628 - no retry when using 'nova image-list' on next configured glance api_servers endpoint
  - Patch thread - Fix retry mechanism for generator results
  - Abandoned - backport downstream Liberty/OSP8 (b/c it merged upstream in stable/liberty)
- Red Hat Bug #1329502 - Inconsistency quota size nova, neutron, openstack command
  - (for pre-triage) If I've understood correctly, when you use neutron rather than nova-network, some statics are incorrect from nova quota-show like floating_ips. There's been a lot of discussion about this, for example:
    - Bug #1263628 - nova floatingip quota conflicts with neutron floating ip quota
    - Bug #1186354 - Limits API doesn't work with Neutron
    - Spec (status: -1 in Mitaka) - Get quota and usage from neutron for limit APIs spec
    - Mailing list thread - proxy quota/limits info from neutron
- Friday downstream bug clean-up:
  - Closed (DUPLICATE) - The instance cannot be deleted if some error occurs in the state of block_device_mapping
  - Closed (DUPLICATE) - when instance fails to launch, attempt to delete it with "nova delete" fails
  - Added missing 'async' so it moves from MODIFIED - When a flavor ID has capital letters, booting a VM will fail with CLI but succeeds with dashboard
  - Closed (WONTFIX) - Inconsistency quota size nova, neutron, openstack command
Week of May 2, 2016
- Bug #1568414 - Database Poison Warnings
- Patch - Retain instance metadata for 'evacuate' on shared storage
- Patch - Fixing a few more ComputeDriver mocks
- Patch - Fix database poison warnings, part 8
- Patch - Trivial: dead code
- Add to calendar: Nova Live Migration Meeting
- Add to calendar: Nova Meeting
- TODO: Watch - Boot strapping Hour
- Storage pools - [notes]
- Coverage from Zuul? Yes! Example: <first-2-digits-of-sha1>/<sha1>/post/nova-coverage-db
- OSP7/Kilo backport - Retain instance metadata for 'evacuate' on shared storage
- Red Hat Bug #1331286 - LibvirtError occurs during evacuation when nfs volume is used
Week of April 25, 2016
- Learned how to do a scratch build
- Watched: Summit keynotes (in the background)
- Testing machine quote from CDW
- Red Hat Bug #1329018 - VMware: Use Port Group and Key in binding details
  - Scratch build (openstack-nova-12.0.3-2.el7.bz1329018)
- Red Hat Bug #1330628 - no retry when using 'nova image-list' on next configured glance api_servers endpoint
  - Patch thread - Fix retry mechanism for generator results
  - backport downstream Liberty/OSP8 (TODO.el7ost)
- Bug #1568414 - Database Poison Warnings
- Patch - Fix database poison warnings, part 5
- Patch - Fix database poison warnings, part 6
- Summit etherpads: https://wiki.openstack.org/wiki/Design_Summit/Newton/Etherpads
- Purple IRC bot: https://anticdent.org/purple-irc-bot.html
- Patch - Wrong mock, wrong mock order
- Patch - Fix database poison warnings, part 7
- Watch summit videos: https://www.youtube.com/user/OpenStackFoundation
- TODO: checkout OSIC (OpenStack Innovation Center)
- Review/co-author: Retain instance metadata for 'evacuate' on shared storage
Week of April 18, 2016
- Bug #1568414 - Database Poison Warnings
- Nova bug clean-up day
- Red Hat Bug #1324238 - When a flavor ID has capital letters, booting a VM will fail with CLI but succeeds with dashboard
  - Patch thread - Revert "nova flavor-show command is inconsistent"
  - originally fixed in master (Liberty/OSP8)
  - backport upstream stable/kilo (unlikely to merge)
  - backport downstream Kilo/OSP7 (python-novaclient-2.23.0-4.el7ost)
- Look into improving the metadata cache - [notes]
- Attempt to backport this to OSP8/Liberty (the code has diverged a bunch) - VMware: Use Port Group and Key in binding details
  - Here's a first pass: http://paste.openstack.org/show/495036/
- Patch - Fix database poison warnings, part 4
- Patch - Trivial: fix mock decorator order
Week of April 11, 2016
- Bug #1568292 - broken review priorities link in devref
- Bug #1568414 - Database Poison Warnings
- Red Hat Bug #1293235 - The instance cannot be deleted if some error occurs in the state of block_device_mapping
- Red Hat Bug #1260426 - when instance fails to launch, attempt to delete it with "nova delete" fails
- Red Hat Bug #1217389 - Faulty status information of OpenStack services after changing the system time
- Attended: OpenStack Toronto: Update on Red Hat's OpenStack Platform 8
- DevStack and non-cirros instances - [notes]
- Patch - Fix database poison warnings, part 2
- Patch - Fix database poison warnings, part 3
- First downstream backports! :)
  - Red Hat Bug #1323654 - Flavor %(id)d extra spec cannot be updated
    - Patch thread - Fix incorrect format string
    - originally fixed in master (Mitaka/OSP9)
    - backport upstream stable/liberty (merged)
    - backport downstream Kilo/OSP7 (openstack-nova-2015.1.3-12.el7ost)
  - Red Hat Bug #1313254 - no retry when using 'nova image-list' on next configured glance api_servers endpoint
    - Patch thread - Fix retry mechanism for generator results
    - originally fixed in master (Newton/OSP10)
    - backport upstream stable/mitaka (merged)
    - backport upstream stable/liberty (not yet merged) <---- !!!!
    - backport downstream Kilo/OSP7 (openstack-nova-2015.1.3-11.el7ost)
  - Red Hat Bug #1318883 - Nova deletes an instance, but doesn't update the database
    - Patch thread - Fix: unable to delete instance when cinder is down
    - originally fixed in master (Newton/OSP10)
    - backport upstream stable/mitaka (merged)
    - backport upstream stable/liberty (merged)
    - backport downstream Kilo/OSP7 (openstack-nova-2015.1.3-13.el7ost)
Week of April 4, 2016
- Do some upstream reviews:
  - https://review.openstack.org/#/c/300751/
- Look into improving the metadata cache - [notes] [more notes]
- Rally: https://wiki.openstack.org/wiki/Rally
- SQLAlchemy Issue #3488 - don't load deferred columns on unexpire for merge with load=False
- Change the number of slow tests to show here (DEFAULT_ROWS_SHOWN)
- I think the unit tests are slower... today vs Feb 5th
- Patch - Fix database poison warnings, part 1
- Red Hat Bug #1324238 - When a flavor ID has capital letters, booting a VM will fail with CLI but succeeds with dashboard
- Backport to Kilo - Revert "nova flavor-show command is inconsistent"
- Review other downstream python-novaclient bugs to see how they are answered etc
- TODO: a trivial downstream backport of ANYTHING ;)
- Red Hat Bug #1323654 - Flavor %(id)d extra spec cannot be updated
- Backport to Liberty - Fix incorrect format string
Week of March 28, 2016
- Red Hat Bug #1318883 - Nova deletes an instance, but doesn't update the database - [notes]
- Red Hat Bug #1291564 - "nova flavor-show" and "nova flavor-delete" can't find the right flavor - [notes]
- Bug #1563547 - Unable to delete instance when cinder is down
- Copy & paste in screen: http://www.saltycrane.com/blog/2008/01/how-to-scroll-in-gnu-screen/#copy-paste
- Hmmm... rejoin-stack.sh is dead: https://review.openstack.org/#/c/291453/
- Watched: What's New in OpenStack Mitaka
- Attended: OpenStack Toronto - Introduction to OpenStack Trove & DBaaS
- TODO - Review history and status of: Unnecessary "SELECT 1" statements
- Red Hat Bug #1322407 - nova reset-password command not able to reset password of instance having selinux in enforcing mode - [notes]
- https://github.com/sdague/failopotamus
- Backport to Mitaka - Fix: unable to delete instance when cinder is down
- Backport to Liberty - Fix: unable to delete instance when cinder is down
Week of March 21, 2016
- Backport to Mitaka - Fix retry mechanism for generator results
- Backport to Liberty - Fix retry mechanism for generator results
- Red Hat Bug #1313254 - no retry when using 'nova image-list' on next configured glance api_servers endpoint
- Patch - Clean up the TestGlanceClientWrapper retry tests
- Debugging (pdb) Tests - https://wiki.openstack.org/wiki/Testr
- Patch - Fix SAWarning in _flavor_get_by_flavor_id_from_db
- Red Hat Bug #1318883 - Nova deletes an instance, but doesn't update the database - [notes]
- Look into improving the metadata cache - [notes]
Week of March 14, 2016
- Back to work!
- Red Hat Bug #1313254 - no retry when using 'nova image-list' on next configured glance api_servers endpoint - [notes]
- Hit - pip install fails when using constraints on keystone
- Bug #1557584 - Broken retry mechanism for 'nova image-list'
- Look into improving the metadata cache - [notes]
Week of February 29, 2016
- Maternity Leave
- Bug #1510234 - Heartbeats stop when time is changed
- Review/testing - Fix Heartbeats stop when time is changed
- Edit OpenStack Python 3 porting blog post
Week of February 15, 2016
- Maternity Leave
- Patch - rpc.init() is being called twice per test
- Patch - WIP: Skip fixtures that need config (where possible)
Week of February 8, 2016
- Maternity Leave
- Patch - Top 100 slow tests: api.openstack.compute.test_versions
- Patch - Top 100 slow tests: api.openstack.compute.test_volumes
- Patch - Top 100 slow tests: api.openstack.compute.test_api
- Patch - Top 100 slow tests: api.openstack.compute.test_image_size
- Patch - Top 100 slow tests: legacy_v2.test_servers
- Patch - Top 100 slow tests: api.openstack.compute.test_flavor*
Week of February 1, 2016
- Maternity Leave
- Patch - Stop using mox (scheduler)
- Patch - Database not needed for test class: PciGetInstanceDevs
- Patch - Database not needed for test class: ConsoleAPITestCase
- Patch - Database not needed for test class: ConductorImportTest
- Patch - Database not needed for RPC serializer tests
- Patch - Database not needed for most crypto tests
- Patch - Database not needed for most nova manage tests
- Patch - Database not needed for most cells messaging tests
- Bug #1533859 - There should be a DB API test that ensures no new tables have soft-delete columns
- Start looking into the top 100 slow tests (proflie)
- Audit tests that hit the database
- Patch babysitting (rebasing, answering code reviews, etc)
- Review & learn from other patches
Week of January 25, 2016
- Maternity Leave
- SQLAlchemy Issue #3629 - SQLite reflection does not preserve primary key constraint names
- Patch babysitting (rebasing, answering code reviews, etc)
Week of January 18, 2016
- Maternity Leave
- Patch - Stop using mox (scheduler)
- Review other "Remove mox patches"
- Bug #1537450 - [gate] functional job fails with TypeError: an integer is required (eventlet)
- Patch - Enable pep8 rule E121
- Patch - Enable pep8 rule E251
Week of January 11, 2016
- Maternity Leave
- Patch - Replace stubs.Set with stub_out (functional tests)
- Patch - Replace stubs.Set with stub_out (fakes)
- Patch - Replace stubs.Set with stub_out (db)
- Patch - Replace stubs.Set with stub_out (os)
- Patch - Replace stubs.Set with stub_out (utils)
- Patch - Replace stubs.Set with stub_out (api)
- Patch - Replace stubs.Set with stub_out (compute api)
- Bug #1532340 - No such file or directory: '/proc/self/mounts'
- Create linux exam study notes (work in progress) & study
- Systems Administration II - Chapter 8
Week of January 4, 2016
- Maternity Leave
- Cull unit test warnings again
- Patch - Remove non-unicode bind param warnings
- Patch - Remove boto deprecation warnings in the unit tests
- Patch - Replace stubs.Set with stub_out (functional tests)
- Patch - Replace stubs.Set with stub_out (fakes)
- Bug #1532340 - No such file or directory: '/proc/self/mounts'
- Create linux exam study notes (work in progress) & study
- Systems Administration II - Chapter 7
Week of November 23, 2015
- Maternity Leave
- Try to start contributing to SQLAlchemy again
- SQLAlchemy Issue #2694 - add autoescape option to startswith, endswith, and contains
- SQLAlchemy Issue #3573 - Postgres: Do not prefix table with schema in: "FOR UPDATE OF <table>"
- Create linux exam study notes (work in progress) & study
Week of November 16, 2015
- Maternity Leave
- Watch: Red Hat Tech Talk - virt-v2v and virt-p2v (Richard Jones)
- Bug #1514863 - Scheduler driver should use Service object
- Patch - Remove IN-predicate warnings
- Bug #1479056 - Fix usage of deprecated methods from oslo.utils.timeutils
- Create linux exam study notes (work in progress) & study
Week of November 9, 2015
- Shadowing database related downstream issues
- Bug #1483872 - RuntimeError: OrderedDict mutated during iteration
- Bug #1514863 - Scheduler driver should use Service object
- Learning to Debug the Gate - notes
- Bug #1359808 - extended_volumes slows down the nova instance list by 40..50%
- Bug #1479056 - Fix usage of deprecated methods from oslo.utils.timeutils
Week of November 2, 2015
- Create linux exam study notes (work in progress)
- 3rd quarter goals
- Updated Stackalytics company affiliation
- Bug #1510234 - Heartbeats stop when time is changed
- Bug #1479056 - Fix usage of deprecated methods from oslo.utils.timeutils
- Bug #1506658 - Loads of unit test failures in Python 3.5: OrderedDict mutated during iteration
- Reviewing upstream patches (with and without actually adding a +1/0/-1)
- Reviewing bugs, looking for ones that are new contributor friendly
- Review the Python 3 porting efforts so far, try to contribute next week
- Reading: https://wiki.openstack.org/wiki/Nova/Mentoring (and its links)
Week of October 26, 2015
- Create linux exam study notes (work in progress)
- Systems Administration II - Chapter 6
- Time warp bug - notes
- Red Hat Bug #1255359 - Flavor cannot be deleted - openstack flavor notes
- Bug #1510234 - Heartbeats stop when time is changed
- Bug #1511487 - 'pause' is not explicitly defined in the hypervisor support matrix
- Bug #1502107 - Method 'get_all' from servicegroup API doesn't implemented in drivers
- Patch - Deprecated: timeutils.strtime & timeutils.isotime
- Patch - Delete dead test code
- Patch - Remove IN-predicate warnings
- Patch - Deprecated: failIf
- Patch - Remove non-unicode bind param warnings
- Patch - Remove "Can't resolve label reference" warnings
- Participate more upstream
Week of October 19, 2015
- Create linux exam study notes (work in progress)
- Systems Administration II - Chapters 1, 2, 3, 4, 5
- Time warp bug - notes
- Red Hat Bug #1255359 - Flavor cannot be deleted - openstack flavor notes
- Bug #1359808 - extended_volumes slows down the nova instance list by 40..50%
Week of October 12, 2015
- Time warp bug - notes
- Red Hat Bug #1255359 - Flavor cannot be deleted - openstack flavor notes
- Bug #1359808 - extended_volumes slows down the nova instance list by 40..50%
- Bug #1503503 - Incorrect argument help for 'nova flavor-create'
- Systems Administration I - Chapters 14, 15
Week of October 5, 2015
- DevStack at a specific version rather than master - notes
- Red Hat Bug #1255359 - Flavor cannot be deleted - openstack flavor notes
- Bug #1503503 - Incorrect argument help for 'nova flavor-create'
- Bug #1359808 - extended_volumes slows down the nova instance list by 40..50%
- Try to help with reviews and bug triage
- Shadow building RHOS RPMs
- Systems Administration I - Chapters 12, 13
Week of September 28, 2015
- Deep dive: What does nova list do under the hood? nova list trace
- Bug #1359808 - extended_volumes slows down the nova instance list by 40..50%
- Red Hat Bug #1255359 - Flavor cannot be deleted - openstack flavor notes
- Systems Administration I - Chapter 11
Week of September 21, 2015
- RDO Liberty - test day notes
- Poor man's query profiler - nova list queries
- Systems Administration I - Chapters 7, 8, 9, 10
Week of September 14, 2015
- Simple DevStack - notes
- Mutli-node DevStack
- Bug #1496135 - libvirt live-migration will not honor destination vcpu_pin_set config
- Systems Administration I - Chapters 4, 5, 6
Week of September 7, 2015
- Bug #1385489 - ResourceTracker._update_usage_from_migrations() is inefficient due to multiple Instance.get_by_uuid() lookups
- Bug clean-up - Low-hanging-fruit
- Systems Administration I - Chapters 1, 2, 3
Older:
- Bug #1441980 - When port group not found in ESXi , get TypeError: 'NoneType' object is not iterable
- Bug #1447528 - vnc configuration should be group 'vnc'
- Patch - Add unit tests for PCI utils
- Patch - Add unit tests for the exact filters

⚠️ GitHub.com Fallback ⚠️