Skip to content
This repository has been archived by the owner on Feb 9, 2024. It is now read-only.

Releases: gravitational/gravity

Release 6.1.41

02 Oct 18:01
37df12c
Compare
Choose a tag to compare

6.1.41 LTS (October 2nd, 2020)

Improvements

  • Improve operation plan sync resiliency in case of temporary etcd issues (#2144).
  • Gravity debug report will now include Gravity resources (#2161).
  • Update a minimum required subnet size to be a /22 (#2183).

Bugfixes

  • Fix an issue with inability to resume operation when gravity-site or etcd are down (#2165).
  • Fix an issue with preflight checks not taking mount overrides into account (#2168).

Release 7.0.17

30 Sep 18:34
8956a96
Compare
Choose a tag to compare

7.0.17 LTS (September 30th, 2020)

Improvements

  • Tune gravity to support up to 1000 nodes (#2159).
  • Introduce reconciliation loop to disable source/dest check when using AWS integrations (#2163).
  • Include additional gravity configuration within debug reports (#2162).
  • Verify agents are active before resuming operations (#2073).
  • Shrink operation resiliency improvements (#2114).
  • Optimize etcd upgrade to avoid steps that are no longer required (#2098).
  • Add check that state dir has expected uid/gid set (#2058).
  • Add check that env variables for http(s) proxy are the correct format (#2062).

Bugfixes

  • Fix an issue with operations stalling when gravity-site or etcd is down (#2166).
  • Fix an issue with wizard installation (#2137).
  • Fix an issue with a cluster reporting as online when some nodes have been partitioned (#2131).
  • Fix an issue with syncing to the registry when using custom planet containers (#2080).
  • Fix an issue with selinux installs rhel/centos 7 (#2012).

Release 6.1.40

25 Sep 23:58
d34e87b
Compare
Choose a tag to compare

6.1.40 LTS (September 25th, 2020)

Improvements

  • Degrade cluster status if any of the nodes are offline (#2130).
  • Make sure upgrade agents are online when resuming or rolling back the operation (#2171).
  • Scale Prometheus/Alertmanager replicas according to the number of cluster nodes (#2145, monitoring#188).

Bugfixes

  • Fix an issue with installer intermittently failing with the "connection refused" error (#2151).
  • Fix an issue with wizard installation (#2135).

Release 6.1.39

18 Sep 20:00
f00c352
Compare
Choose a tag to compare

6.1.39 LTS (September 15th, 2020)

Bugfixes

  • Fix a regression in tele build when tele would fail to pull the dependent planet package (#2126).

Release 6.1.38

12 Sep 00:40
5f105a2
Compare
Choose a tag to compare

6.1.38 LTS (September 11th, 2020)

Improvements

  • Increase default LimitNOFile parameter within planet's systemd to 655350 (#2092, planet#743).
  • Upgrade Helm to v2.15.2 (#2118, planet#747).
  • Shrink operation resiliency improvements (#2096).

Bugfixes

  • Fix an issue with etcd not being upgraded on the worker nodes (#2097, planet#740).
  • Fix an issue with install not being able to uninstall existing failed planet service (#2110).
  • Fix an issue with upgrade when custom planet container is used (#2094).

Release 5.5.52

09 Sep 00:42
de2f950
Compare
Choose a tag to compare

5.5.52 LTS (September 8th, 2020)

Improvements

  • Make sure that agents are available before resuming or rolling back an upgrade (#2044).
  • Restart flannel before upgrade from older versions to workaround potential overlay issues (#2043).
  • Remove the requirement for all nodes to have the same operating system version from preflight checks (#2048).
  • Display active Kapacitor alerts in gravity status (#2039).
  • Update gravity status to display a warning upon state directory ownership change (#2034, planet#727, satellite#255).
  • Validate HTTP_PROXY and HTTPS_PROXY variables set via RuntimeEnvironment resource (#2049).
  • Update the log collector rotation policy to keep ten 256MB log files (#2033, logging#70).
  • Drop excessive metrics from InfluxDB during an upgrade (#2091, monitoring#187).

Bugfixes

  • Fix an issue with remote_syslog process leaking CPU time (#2084, logging#71).
  • Fix an issue with etcd not being upgraded on the worker nodes (#2086, planet#738).
  • Fix an issue with installation sometimes failing when trying to install a cluster license (#2053).
  • Fix an issue with planet agent connection to serf periodically hanging after cluster networking issues (#2079, planet#739, satellite#262).

Release 7.0.16

31 Aug 21:14
87cfb76
Compare
Choose a tag to compare

7.0.16 LTS (August 31st, 2020)

Bugfixes

  • Fix an issue with disconnected Teleport nodes when rotating all masters within the cluster (#2029).
  • Fix an issue with installation sometimes failing when trying to install a cluster license (#2037).
  • Fix an issue with priority class incorrectly being set on the hook jobs (#2066).

Release 6.1.37

31 Aug 21:22
b03d314
Compare
Choose a tag to compare

6.1.37 LTS (August 31st, 2020)

Improvements

  • Add checker that validates Gravity data directory user/group ownership (#2057).
  • Validate HTTP_PROXY and HTTPS_PROXY variables set via RuntimeEnvironment resource (#2061).

Bugfixes

  • Fix an issue with installation sometimes failing when trying to install a cluster license (#2054).

Release 6.1.36

20 Aug 16:03
9385abb
Compare
Choose a tag to compare

6.1.36 LTS (August 20th, 2020)

Bugfixes

  • Fix an issue with upgrading applications that use custom planet containers that contained corrupted version metadata (#2032).
  • Fix an issue with upgrades if etcd is deployed within kubernetes and scheduled to a master (#2024, planet#721).
  • Fix an issue with corrupted version metadata within custom planet containers (#2024, planet#717).

Release 7.0.15

20 Aug 17:24
2f61b2f
Compare
Choose a tag to compare

7.0.15 LTS (August 19th, 2020)

Improvements

  • Increase default LimitNOFile parameter within planet's systemd to 655350 (#2027, planet#725).

Bugfixes

  • Fix an issue with DNS application priority class not having any effect (#1989).