github/kubernetes
4
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
119,149 Commits 1 Branch 31 Tags
ef05f59448ecaac6e1330fac6af76b5c43e26d2a
Commit Graph

944 Commits

Author SHA1 Message Date
Marian Lobur
5d80d6e7c3 Make cpu request of kube proxy configurable by env variable. 2021-07-02 16:00:56 +02:00
Piotr Tabor
de442ef860 Retry hostname->IP: [Errno -2] Name or service not known 
During cluster configuration, the hostname is getting resolved to IP,
as etcd requires IP address as listening address.

Due to connectivity flakes or delayed network inititalization, sometimes
the IP fails to be resolved to a name with following error:
```
[Errno -2] Name or service not known
```
that leads to attempt to run etcd with empty flag.

The PR adds a proper retry (up to 5 minutes) in case the connectivity
problems happens.

I considered alternatives like: `getent hosts foo`, but unfortunetelly thay
can return IPv6 that etcd is not ready for (yet).
 2021-07-01 12:20:07 +02:00
Vinayak Goyal
774d228637 remove the path if it exists before writing pki data. 
if setfacl is called before chmod g+r at anypoint during the lifetime of
the cluster then the default group does not have read permissions on the
file. so we explicitly grant the default group read permissions. See
https://gist.github.com/mmdriley/85ca34f711acbec4b1b94902add488e5 for a
repro.
 2021-06-18 11:03:37 -07:00
Joseph Anttila Hall
9d514b2de4 Konnectivity: tune flags for larger clusters (5k nodes). 2021-06-10 14:05:44 -07:00
Kubernetes Prow Robot
9d27400fe2 Merge pull request #102040 from njuptlzf/fix_conversion 
Fix auditing failed of request: encoding failed
 2021-06-05 19:58:38 -07:00
njuptlzf
7b0fbb7292 add audit log test cases for cross-group subresource 2021-06-06 09:52:05 +08:00
Marek Siarkowicz
4ebc0c94a4 Remove legacy metrics client from podautoscaler 2021-06-04 23:06:32 +02:00
Kubernetes Prow Robot
d541872f9a Merge pull request #102239 from Haleygo/clean-up-AlgorithmProvider-flag-and-pkg 
clean up algorithmprovider pkg and remove scheduler deprecated algorithm-provider flag
 2021-05-27 00:54:23 -07:00
Haleygo
2769e99dba remove scheduler deprecated algorithm-provider flag and clean up algorithmprovider pkg 2021-05-26 13:19:44 +08:00
Kubernetes Prow Robot
06d44d2f42 Merge pull request #101168 from mikedanese/warning 
add a warning about the filter table
 2021-05-24 21:48:40 -07:00
Kubernetes Prow Robot
77937b1e8e Merge pull request #101628 from bobbypage/addon-termination-handler 
Remove node termination handler addon
 2021-05-24 11:31:39 -07:00
Kubernetes Prow Robot
e8cf412e5e Merge pull request #101881 from vinayakankugoyal/konnectivity 
Update konnectivity network proxy server to run as non-root, by defau…
 2021-05-13 23:16:04 -07:00
Vinayak Goyal
b951b9349f Update konnectivity network proxy server to run as non-root, by default in kube-up. 2021-05-13 12:35:34 -07:00
Avritt Rohwer
0a5a697882 Fix bug in retry-forever usage. 
- Push retry-forever wrapping to curl invocations.
- Collect curl retry flags into a single variable.
- Remove 'sudo: false' in master.yaml, is unnecessary and breaks older
  cloud-init versions.
- Change log-error status reason to be more accurate.
- Fix the some 'python' invocations to 'python3'.
 2021-05-12 09:22:20 -07:00
Kubernetes Prow Robot
ca0c04e4d3 Merge pull request #101164 from vinayakankugoyal/apiservernonroot 
Run control-plane as non root in kube-up.
 2021-05-06 17:33:14 -07:00
Kubernetes Prow Robot
8955f55fcf Merge pull request #101678 from vinayakankugoyal/goodbye-basicauth 
Remove remnants of basic auth from cluster bootstrap.
 2021-05-06 14:14:14 -07:00
Vinayak Goyal
6aa495ddc6 Revert - Recursive chown the /etc/srv/sshproxy if kube-apiserver is running as non root. This way if a key already exists we will be able to read it. 2021-05-06 14:02:53 -07:00
David Porter
dac06aefb0 Revert "Revert "cluster: Use python3 everywhere"" 
This reverts commit 7038338e0f.
 2021-05-03 21:43:15 -07:00
Kubernetes Prow Robot
c5b900b69c Merge pull request #97399 from davidxia/comment-typo 
Fix typo in comment
 2021-05-01 04:57:59 -07:00
Vinayak Goyal
b87762966d Remove remnants of basic auth from cluster bootstrap. 2021-04-30 11:23:14 -07:00
David Porter
e02ff0687e Remove node termination handler addon 2021-04-29 14:42:23 -07:00
Paco Xu
7038338e0f Revert "cluster: Use python3 everywhere" 2021-04-26 11:21:44 +08:00
David Porter
3f87f4f278 Use python3 everywhere 2021-04-23 14:33:58 -07:00
Kubernetes Prow Robot
ae35c6f10c Merge pull request #101255 from basantsa1989/stack-type 
Adding stack-type to gce cloud config (to be used for dual stack in legacy-cloud-providers gce code)
 2021-04-22 15:55:28 -07:00
Basant Amarkhed
e15d811652 Adding stack-type to cloud config (to be used for dual stack in legacy-cloud-providers code) 2021-04-19 19:06:55 +00:00
Shihang Zhang
297ad30610 dnat to 169.254.169.252 for metadata server traffic 2021-04-19 10:47:51 -07:00
Kubernetes Prow Robot
28c877dcb6 Merge pull request #101043 from benhxy/tls-2 
Use GKE specific configuration for kubeconfig file generation
 2021-04-16 11:54:51 -07:00
Mike Danese
ba3fc65072 add a warning about the filter table 2021-04-15 16:22:28 -07:00
Maciej Borsz
493adbada9 Do not grep for curl --help for --retry-connrefused 2021-04-14 08:32:21 +02:00
Kubernetes Prow Robot
f1c037889d Merge pull request #100770 from avrittrohwer/configure-script-logging 
Add configure script logging instrumentation
 2021-04-13 18:06:42 -07:00
Kubernetes Prow Robot
318db993c8 Merge pull request #101020 from cindy52/bugfix/etcd 
Change file owner of /mnt/disks/master-pd/var/etcd  instead of /var/etcd
 2021-04-13 12:09:47 -07:00
Avritt Rohwer
d4495183c9 Add configure script logging instrumentation. 
- Add log functions to facilitate debug logging.
- Wrap commands called in main with debug logging.
- Configure a systemd service to forward the logs to the serial port.
- Add a 'retry-forever' function to harden download steps.
- Add default value support to 'get-metadata-value' function.
- Fix some spellcheck lints.
 2021-04-13 09:30:49 -07:00
Ben Hu
e3270e532c GKE specific kubeconfig 2021-04-12 22:47:39 +00:00
Cindy Guo
03f60f4b60 chown on /mnt/disks/master-pd/var/etcd instead of /var/etcd 2021-04-12 08:21:01 +00:00
Antonio Ojea
93f4727aab gce configure containerd default_runtime_name 
move config to v2
 2021-04-11 00:48:22 +02:00
Cindy Guo
9f058079d2 run etcd as nonroot 
Co-authored-by: Vinayak Goyal <vinayakankugoyal@gmail.com>
 2021-04-08 20:51:45 +00:00
varsha teratipally
90983f66e4 Moving docker options to daemon.json 
As per the new docker guidelines about customizing the options
like adding registry-mirrors, moving the options to daemon.json
 2021-03-10 19:14:48 +00:00
Kubernetes Prow Robot
874877fa44 Merge pull request #99216 from ruiwen-zhao/remove_modprobe 
Remove modprobe configs from configure-helper
 2021-02-22 17:24:32 -08:00
Cong Liu
03709c0ece Add arm64 support for GCE node configuration 
Fix typo

Add TODO
 2021-02-19 14:22:26 -08:00
ruiwen-zhao
c053b232ba Remove modprobe configs from configure-helper 2021-02-18 22:57:44 +00:00
Benjamin Elder
299c561b10 portably configure tempdir in configure-helper.sh 
fixes a `make test` failure on macOS
 2021-02-12 01:15:14 -08:00
Matthew Cary
9a7dcd36c1 Disallow local loopback for volume hosts 
Change-Id: Ic356c3f859057153cfad97327f1938792a1a512c
 2021-01-26 17:12:51 -08:00
Kubernetes Prow Robot
1a67280508 Merge pull request #98037 from vinayakankugoyal/kube-controller-manager-crp 
Update configure-helper.sh to early exit from start-kube-controller-m…
 2021-01-25 12:38:59 -08:00
Vinayak Goyal
31807032e0 Update configure-helper.sh to early exit from start-kube-controller-manager if kube-controller-manager is deployed through CRP. 2021-01-20 16:22:46 -08:00
Michael Taufen
9f9e235b9d Mount /var/lib/kubelet/pki on tmpfs 
This helps avoid some rare instances of corrupt cert files
that cause Kubelet to crash-loop after node reboots, e.g.
if Kubelet opens the file during the shutdown but is unable
to write it.
 2021-01-08 18:04:35 -08:00
Jian Zeng
8c1971e17c chore(gce): pass auth flags to KCM and KS 
Pass flags `--authentication-kubeconfig` and
`--authorization-kubeconfig` to controller-manager and scheduler,
so that we could grab metrics from their secure ports in tests.
 2021-01-06 12:56:39 +08:00
Sergey Kanzhelev
d78db9f161 configure docker on containerd nodes so it wouldn't reserver 172.17 subnet 2020-12-23 18:49:57 +00:00
David Xia
0756e54dfc Fix typo in comment 2020-12-21 20:02:20 -05:00
Ben Hu
9581c40887 Revert "Use host IP instead of localhost for control plane component kubeconfig files." 
This reverts commit 49afcfa5f2.
 2020-12-11 22:36:39 +00:00
Maciej Borsz
7f09d59215 Migrate etcd's livenessProbe to etcdctl endpoint health. 
Change-Id: Ie19c844050c75e3d1c4b431d09ba0ac851c5317b
 2020-12-11 12:43:02 +01:00