rajashree
1ecf6effbf
Reconcile workerplane for NotReady control hosts
2020-03-20 13:37:37 -07:00
rajashree
fc3709507d
Reset error to nil if lister works on retries
2020-03-13 15:16:25 -07:00
rajashree
6b25bcf3e0
Remove ignore-upgrade label from zero downtime upgrade
2020-03-10 23:36:17 -07:00
rajashree
bb6873ce48
Addresses following issues:
...
1. Compare maxUnavailable with powered off hosts before attempting to reconcile
NotReady hosts
2. Include powered off hosts as failed hosts for controlplane upgrade to return error
3. Change coredns upgrade strategy. With addons changes it was changed to have the k8s
default value for a deployment of 25% maxUnavailable and maxSurge. This commit changes it
back to maxUnavailable of 1 to avoid dns addon upgrade issues
2020-03-07 14:22:35 -08:00
rajashree
c61d531af1
Rename ignore label and return error on controlplane failure
2020-03-04 15:20:48 -08:00
rajashree
115c11b864
Fix maxunavailable reset
2020-02-28 13:56:19 -08:00
rajashree
e27a05f8b1
Attempt upgrade on NotReady hosts
2020-02-26 14:14:42 -08:00
rajashree
fd1e1289de
Return correct error if maxUnavailable hit for control
2020-02-24 17:56:02 -08:00
rajashree
968a399f26
Accept label to ignore nodes during upgrade
...
RKE does a cluster scan to find the unreachable hosts, and if that number
is same as or exceeds maxUnavailable, upgrade won't proceed.
This commit introduces a label users can provide for their nodes so they
don't get counted as unavailable and are excluded from upgrade.
This commit also includes a couple of bug fixes
2020-02-22 14:23:47 -08:00
Rajashree Mandaogane
92714e5523
Merge pull request #1800 from mrajashree/workers_upgrade
...
Change RKE upgrade logic for zero downtime
2020-02-06 11:03:29 -08:00
rajashree
11678a3f98
Change RKE upgrade logic for zero downtime
2020-02-05 16:19:47 -08:00
Sebastiaan van Steenis
23e98603bd
Determine etcd s3 snapshots by s3 config presence
2020-01-20 18:43:35 +01:00
Alena Prokharchyk
fb90d1429c
Permissions fix for etcd restore with custom uid/gid
2019-12-06 10:22:28 -08:00
Alena Prokharchyk
2bc68c7118
Correct system image for etcd snapshot removal
...
when removal is executed as a part of restoration
2019-12-03 15:58:31 -08:00
moelsayed
2006a61712
Move snapshot remove logic to rke-tools
2019-11-11 16:03:44 -08:00
Alena Prokharchyk
fa9410faeb
Return proper error in etcd healthcheck
2019-11-04 15:10:05 -08:00
Sebastiaan van Steenis
adc5941fd9
Add per node kubelet server certificate
2019-10-31 15:56:44 -07:00
moelsayed
372393ac1b
Add Secret Encryption Provider Support
2019-10-29 14:10:32 -07:00
Dax McDonald
ad678b6a32
Remove dead code
2019-10-04 15:02:14 -07:00
Dax McDonald
8022b815b3
Remove uneeded nil check
2019-10-04 15:01:53 -07:00
Sebastiaan van Steenis
4a4a2357d3
Do not randomize etcd-servers in kube-apiserver
2019-09-04 15:52:02 -07:00
Sebastiaan van Steenis
32bf922fc5
Apply MCS label when privileged is false
2019-08-30 10:42:11 -07:00
Sebastiaan van Steenis
ac16bd8b33
Configure MCS labels if selinux is enabled
2019-08-22 13:45:04 -07:00
Sebastiaan van Steenis
5d8d538ce0
Sync log messages for etcd snapshot containers
2019-08-22 13:30:51 -07:00
Frank Mai
277797df0f
Support to generate Windows worker plan
...
- Put Windows worker plan generating back to reduce the changing from
Windows on rancher/rancher
- Prepare for rke bootstraps Windows cluster
**Issue:**
https://github.com/rancher/rancher/issues/16460
2019-08-21 20:50:31 -07:00
Sebastiaan van Steenis
65df510251
Improve etcd snapshot logging
2019-08-21 20:40:40 -07:00
moelsayed
06e87ebabb
Remove uncompressed snapshot after restore
2019-08-20 12:50:25 -07:00
Darren Shepherd
f8bac2c059
Update to new certs package since latest k8s dropped it
2019-08-19 11:02:43 -07:00
Sebastiaan van Steenis
3f94e86706
Revert "Add per node kubelet server certificate"
...
This reverts commit b860e634db
2019-08-08 09:49:47 -07:00
Sebastiaan van Steenis
b860e634db
Add per node kubelet server certificate
2019-07-31 14:54:43 -07:00
moelsayed
688d4aedd7
support etcd custom uid/gid
2019-07-29 17:24:16 -07:00
Sebastiaan van Steenis
2b53fefbee
Add parameter for s3 folder
2019-07-19 13:25:32 -07:00
kinarashah
217e1b41b8
generate correct default rketools
...
always use rke's default k8s's rke-tools, even if rancher's default k8s
changes. This is based on assumption that change in rke-tools would also
require a new rke version.
2019-07-18 14:48:48 -07:00
galal-hussein
55b1b4db7c
use healthz endpoint for kubelet healthcheck
2019-07-18 14:22:07 -07:00
moelsayed
d6b66eb8f7
fix md5sum command
2019-07-17 17:11:54 -07:00
moelsayed
17320083e9
Use etcd service extra_env in backup containers
2019-07-17 16:42:26 -07:00
moelsayed
a61a4acc4e
remove rolling snapshot container during clean up
2019-07-11 14:27:55 -07:00
moelsayed
058f196e72
Fix worker/controlplane reconcile logic
2019-07-11 14:27:55 -07:00
Sebastiaan van Steenis
173669c8a5
Add retry logic to Docker interactions
2019-06-26 12:15:32 -07:00
moelsayed
2c907f9f21
rename EndpointCA
2019-06-25 14:17:53 -07:00
moelsayed
38c31b9766
Add option to pass custom CA certificate for S3 backend
2019-06-20 15:00:00 -07:00
Sebastiaan van Steenis
867c0f44e6
Add zip extension to compressed snapshot
2019-06-12 10:06:02 -07:00
kinarashah
1a1080a234
always use DefaultRKETools for etcd snapshot
2019-06-11 12:52:25 -07:00
Sebastiaan van Steenis
bc09408581
Add etcd snapshot compression
2019-06-07 15:17:22 -07:00
Sebastiaan van Steenis
28d907281f
Dont restart s3 snapshot download container
2019-06-05 13:40:29 -07:00
Erik Wilson
581e3389c4
Reorder etcd servers list
2019-05-28 09:50:29 -07:00
galal-hussein
c89198b9ff
Add :z to backup volume bind mounts
2019-04-08 09:53:50 -07:00
moelsayed
06b709e888
Add RemoveEtcdSnapshot
2019-04-05 13:51:08 -07:00
moelsayed
066f9300a8
Wait for backup server to start before status check
2019-03-16 18:49:25 -07:00
moelsayed
bb2b5a4f36
Clean old containers before etcdbackup operations
2019-03-15 13:28:00 -07:00
