0.69.26 (2018-01-23)
Merged pull requests:
- 1st iteration of paasta spark-run #1709 (huadongliu)
v0.69.24 (2018-01-19)
Merged pull requests:
v0.69.23 (2018-01-18)
Closed issues:
- Old Marathon apps can fail to be cleaned up (stuck bounces) #1706
Merged pull requests:
- Make paasta_tools.bounce_lib.kill_old_ids raise exceptions for deployd #1707 (nhandler)
- Rename the bespoke-autoscale command to just 'autoscale' #1705 (solarkennedy)
- Always emit 'instances' to signalfx for knowing what the desired instance count is #1704 (solarkennedy)
- Add missing space #1702 (nhandler)
- Tweaks to paasta rightsizer script. #1701 (thebostik)
v0.69.22 (2018-01-15)
Merged pull requests:
- Check marathon replication on non-default reg #1700 (mattmb)
- Tried to clarify the paasta stop message again #1691 (solarkennedy)
v0.69.21 (2018-01-12)
Merged pull requests:
- Fix randomly closing ZK connections in paasta-api #1699 (mattmb)
- validate error message improvement #1698 (chlgit)
v0.69.20 (2018-01-10)
Merged pull requests:
- Close connections to Zookeeper properly #1697 (keymone)
- Revert the Mesos CPU sanity check and added Filter mesos cpu value #1690 (matfra)
- Enabled paasta_remote_run to connect to diffferent mesos clusters #1680 (huadongliu)
v0.69.19 (2018-01-10)
Merged pull requests:
v0.69.18 (2018-01-10)
Merged pull requests:
- When calling deploy_marathon_service from deployd, only fetch apps from relevant marathons. #1692 (EvanKrall)
- Use marathon 1.4.10 in itests #1689 (nhandler)
- Added a script to emit additional paasta metrics #1674 (solarkennedy)
- Added a bespoke-autoscale cli helper #1671 (solarkennedy)
v0.69.17 (2018-01-02)
Merged pull requests:
v0.69.16 (2017-12-24)
Merged pull requests:
- If you don't specify a default value, then dict.pop KeyErrors if the key doesn't exist. #1687 (EvanKrall)
- Fixed override flag on paasta_cluster_boost and print status for any action #1686 (matfra)
v0.69.15 (2017-12-22)
Merged pull requests:
- Pop stats from cache instead of del'ing, so we don't get a KeyError if the stats aren't cached. #1684 (EvanKrall)
- Perform paasta status ssh in parallel #1679 (henryzhangsta)
v0.69.14 (2017-12-22)
Merged pull requests:
- Bumping version of docker-compose in tox #1683 (matfra)
- Allow auto hostname unique in non-prod clusters #1682 (nhandler)
- Ignore all .cache directories #1681 (henryzhangsta)
- Wait for deployment refactoring #1676 (oktopuz)
v0.69.13 (2017-12-22)
Merged pull requests:
v0.69.12 (2017-12-21)
Merged pull requests:
- Clear stats cache before fetching a second time #1678 (EvanKrall)
- Remove extraneous
\# noqa
comments. #1675 (asottile)
v0.69.11 (2017-12-21)
Merged pull requests:
v0.69.10 (2017-12-20)
Closed issues:
- Delete Me #1673
Merged pull requests:
- help wanted -- fix issue 842 #1670 (white105)
- adds a try/except to update_soa_memcpu #1667 (fboxwala)
- Skip wait-for-deployment if mark-for-deployment failed #1666 (oktopuz)
- adds -p to update_soa review branch command #1665 (fboxwala)
- Fix tail discovery, date handling and upgrade yelp_clog #1662 (fede1024)
- RFC: Automatically add ['hostname', 'UNIQUE'] to small services #1652 (nhandler)
v0.69.9 (2017-12-15)
Merged pull requests:
- Avoid using default event loop on secondary threads. #1664 (EvanKrall)
- Perfscript #1663 (fboxwala)
- Make setup_marathon_job ignore instances starting with _ #1655 (nhandler)
v0.69.8 (2017-12-14)
Merged pull requests:
- PAASTA-13596: fix local_run with healthcheck_only #1659 (oktopuz)
- Reexec paasta local-run --pull with sudo when the effective uid is not 0 #1656 (oktopuz)
- Convert drain_lib to asyncio so we can drain in parallel. #1653 (EvanKrall)
v0.69.7 (2017-12-12)
Merged pull requests:
- bugfixes for pause_service_autoscaler #1658 (fboxwala)
- Passing boost arguments only when action is set #1657 (matfra)
- deletes pid decision code #1654 (fboxwala)
- Add remedy commands to the check_mesos_outdated_tasks output #1651 (oktopuz)
v0.69.6 (2017-12-07)
Merged pull requests:
- Perfscript #1649 (fboxwala)
- Make metastatus able to group by multiple values properly #1642 (EvanKrall)
- pytest.fixture for SystemPaastaConfig #1641 (oktopuz)
v0.69.5 (2017-12-06)
Merged pull requests:
- Fixes a small typo in the autoscaling documentation #1650 (fhats)
- A few python3 things #1648 (asottile)
- Make marathon_dashboard work for non-sharded clusters #1647 (nhandler)
- add variables to string interpolation #1645 (Rob-Johnson)
v0.69.4 (2017-12-05)
Merged pull requests:
- Fix mypy for cluster_boost #1646 (nhandler)
- Initial Support for a Sharded Marathon Dashboard #1601 (nhandler)
v0.69.3 (2017-12-04)
Merged pull requests:
- v0.69.2 #1644 (oktopuz)
- paasta local-run --no-healthcheck shouldn't publish a container's ports #1643 (oktopuz)
- Adding Cluster Boost feature (ready to merge) #1587 (matfra)
v0.69.1 (2017-12-01)
Merged pull requests:
- Fix paused autoscaler logic bug #1640 (solarkennedy)
v0.69.0 (2017-11-30)
Merged pull requests:
- Convert cli/cmds/sysdig.py to sharded marathon and clean up unsharded code/test config. #1628 (EvanKrall)
v0.68.12 (2017-11-29)
Merged pull requests:
- Pass MarathonClients, not list of MarathonClient, to get_autoscaling_info in status_marathon_job_verbose. #1639 (EvanKrall)
- Don't docker-compose pull in example_cluster; see bc3029fb3f01eadb931105eea480c9ec21c18ea8 #1638 (EvanKrall)
- Show marathon dashboard links in paasta status #1636 (oktopuz)
- Tweak mypy settings to allow for more granular type checking #1627 (EvanKrall)
v0.68.11 (2017-11-28)
v0.68.10 (2017-11-27)
Merged pull requests:
v0.68.9 (2017-11-23)
Closed issues:
- Cluster autoscaler gives up if too many slaves aren't found when scaling down cancelled_running resource #1619
Merged pull requests:
- Fix TypeError in status_marathon_job_verbose #1635 (oktopuz)
- fix bug in autoscaler for sharded marathon #1634 (Rob-Johnson)
- Support arrays in dashboard_links.json #1633 (oktopuz)
- Allows the cluster autoscaler to act on cancelled_running sfrs with missing instances #1620 (matthewbentley)
- Always fix the slack of old marathon apps when bouncing #1605 (solarkennedy)
- deprecates PID policy #1579 (fboxwala)
v0.68.8 (2017-11-20)
Merged pull requests:
- P1: Fix metrics names #1631 (matthewbentley)
- Always consider num_at_risk_tasks when calculating instance_counts du… #1623 (solarkennedy)
v0.68.7 (2017-11-20)
Closed issues:
- RFC: Emit metrics from cluster autoscaler #1600
Merged pull requests:
- Reuse paasta_itest images #1625 (oktopuz)
- Move deployd metrics to be generic #1621 (matthewbentley)
- Emit metrics from cluster autoscaler #1614 (matthewbentley)
- Adjust check_mesos_duplicate_frameworks for sharding #1609 (oktopuz)
v0.68.6 (2017-11-14)
Merged pull requests:
- Fix silly bug in service autoscaler where I was passing a MarathonClients object instead of a list of MarathonClient objects. #1626 (EvanKrall)
v0.68.5 (2017-11-13)
Merged pull requests:
- Updated task-processing to 0.0.5 #1624 (huadongliu)
- Make service autoscaler work with multiple marathon shards. #1618 (EvanKrall)
v0.68.4 (2017-11-10)
Merged pull requests:
- Make PaaSTA API work with sharded marathon #1617 (oktopuz)
- RFC: Add cluster autoscaler docs #1616 (matthewbentley)
- Make "paasta local-run" try to use the same port repeatedly #1615 (chriskuehl)
- make graceful-app-drain using marathon sharded code #1613 (oktopuz)
- Make paasta local-run lock on pulls #1608 (solarkennedy)
- Add bounce logic for json secret files #1607 (mattmb)
- skip unreachable tasks when draining #1595 (somic)
- Added type annotations to generate_deployments_for_service #1594 (solarkennedy)
- Make check_marathon_services_replication work with sharded marathon #1590 (EvanKrall)
v0.68.3 (2017-11-08)
Merged pull requests:
- increase read timeout for maintenance api calls #1612 (somic)
- Make paasta_get_num_deployments using marathon sharded code #1611 (oktopuz)
- delete_old_marathon_deployments to use marathon sharded code #1610 (oktopuz)
- Cluster autoscaler typing #1606 (matthewbentley)
- Fixed paasta_remote_run and added task logging executor #1603 (huadongliu)
v0.68.2 (2017-11-06)
Closed issues:
- make autoscaling_lib unit tests not sleep #1589
Merged pull requests:
- extract_args doesn't return a constraint #1602 (Rob-Johnson)
- Remove unnessesary calls to sleep in test_autoscaling_cluster_lib #1599 (matthewbentley)
- paasta status: fix string interpolation in error message #1598 (chriskuehl)
- check_mesos_outdated_tasks #1596 (oktopuz)
- [cluster autoscaler] expand logic around scaling cancelled resources #1584 (matthewbentley)
v0.68.1 (2017-11-02)
Merged pull requests:
- Pass a list of clients, not a MarathonClients object, to get_marathon_apps_with_clients in get_at_risk_service_instances. Add type annotations to deployd.common and deployd.watchers and their tests. #1597 (EvanKrall)
- allow specifying pool in adhoc schema #1593 (Rob-Johnson)
- Fix autoscaling info for metastatus #1592 (matthewbentley)
- Fix error when a pool isn't ready when the cluster autoscaler tries to run #1591 (matthewbentley)
- uses meteorite to push autoscale events to signalfx #1588 (fboxwala)
- set the pool correctly in remote-run launches #1577 (Rob-Johnson)
* This Change Log was automatically generated by github_changelog_generator