568 Commits

Author SHA1 Message Date
Thomas Schubart
8c6c0f5a8b [ws-manager] Update golden files 2023-02-14 11:41:19 +01:00
Thomas Schubart
70e7cac44f [ws-manager] Do not schedule on experimental nodes 2023-02-14 11:41:19 +01:00
Pudong Zheng
377f00cb02 [ws-manager] Add custom closed timeout
Co-authored-by: Wouter Verlaek <wouter@gitpod.io>
2023-02-10 11:16:15 +01:00
Wouter Verlaek
cb192728d1 Update Go dependencies 2023-02-08 16:47:46 +01:00
Manuel Alejandro de Brito Fontes
bcdde44633 Switch from alpine to wolfi 2023-02-05 12:10:43 +01:00
Manuel Alejandro de Brito Fontes
89eb6f1603 Update tests 2023-01-30 16:36:39 +01:00
Manuel Alejandro de Brito Fontes
92e11940d0 Do not try to connect to not ready ws-daemon pods 2023-01-30 16:36:39 +01:00
Christian Weichel
abeea17ab5 Move ws-manager-mk2 CRD to API package 2023-01-23 14:14:30 +01:00
utam0k
3f352c3748 ws-manager: Add a log and trace not to overlook workspace schedule failure 2023-01-19 11:52:27 +01:00
JenTing Hsiao
8bd4df75e7 ws-manager: skip failure from terminated message
For some reason, the pod is killed with unknown container status and no taints on the underlying node.
Therefore, we skip extracting the failure from the terminated message.

Signed-off-by: JenTing Hsiao <hsiaoairplane@gmail.com>
2023-01-13 06:06:52 +01:00
JenTing Hsiao
9f33031a6d ws-manager: add log to help troubleshoot
Signed-off-by: JenTing Hsiao <hsiaoairplane@gmail.com>
2023-01-10 08:22:49 +01:00
JenTing Hsiao
40c1d2eacc ws-manager: avoid duplicate logs
Signed-off-by: JenTing Hsiao <hsiaoairplane@gmail.com>
2023-01-09 10:50:48 +01:00
Pavel Tumik
fda0234ae2 fix headless and prebuild pods not removing never ready annotation 2023-01-06 17:28:46 +01:00
JenTing Hsiao
316dbdb52a ws-manager: add new reason out-of-space for workspace_stops_total metric
Signed-off-by: JenTing Hsiao <hsiaoairplane@gmail.com>
2023-01-05 11:27:44 +01:00
JenTing Hsiao
bb96008cb6 Fix the metric workspace_starts_failure_total and workspace_stops_total
Increase the metric workspace_starts_failure_total when the workspace stops and
the pod's annotation "gitpod/never-ready" exist. It means the workspace pod never
went into the running state, probably because
- the workspace pod can't allocate to the workspace node
- content initialization failure
- the workspace pod is never ready

Increase the metric workspace_stop_total when the workspace stops and
the pod's annotation "gitpod/never-ready" does not exist. So, the metric
workspace_stop_total does not increase on the start failure.

Signed-off-by: JenTing Hsiao <hsiaoairplane@gmail.com>
2023-01-04 23:47:44 +01:00
JenTing Hsiao
4af9af4a1e Apply suggestions from code review
Co-authored-by: Pavel Tumik <18602811+sagor999@users.noreply.github.com>
2023-01-03 23:48:43 +01:00
JenTing Hsiao
af04243b3b If dryRun, return gRPC FailedPrecondition
If dryRun enables on StartWorkspace/StopWorkspace/TakeSnapshot,
return gRPC FailedPrecondition status code and error message
`under maintenance mode`.

Signed-off-by: JenTing Hsiao <hsiaoairplane@gmail.com>
2023-01-03 23:48:43 +01:00
JenTing Hsiao
a10dc221fb Watch the config.manager change
Signed-off-by: JenTing Hsiao <hsiaoairplane@gmail.com>
2023-01-03 23:48:43 +01:00
Pavel Tumik
edbaaae0c5 [ws-manager] remove unneeded metric 2023-01-03 19:37:43 +01:00
utam0k
24a7d195db ws-manager: Update and add golden files 2022-12-28 23:23:37 +01:00
utam0k
a7fe67dd17 ws-manager: Prevent to slip the CREATING phase 2022-12-28 23:23:37 +01:00
utam0k
67d810e226 ws-manager: Improve the peformance of reconcile handlerr 2022-12-28 23:23:37 +01:00
utam0k
c40fcc8ae6 ws-manager: Prevent to record duplicated metric 2022-12-20 13:29:29 -03:00
Wouter Verlaek
72d9152a71 [image-builder] Enable TLS in workspace clusters
Fix cyclic import cycle

Update altnames

Update golden testdata, todo

Update grpc opts

Testing

Remove blocking dial

Only add TLS in ws cluster

Conditional TLS

Add comments
2022-12-20 02:45:29 -03:00
utam0k
9e22744f98 ws-manager: A bit improve the code not to use the magic number 2022-12-20 00:41:29 -03:00
Manuel Alejandro de Brito Fontes
f22a92a72f Wait for the workspace secret before creating the workspace pod 2022-12-15 21:36:25 -03:00
Pavel Tumik
df1428d6f8 [wsmanager] improve error tracing in finalize workspace content 2022-12-11 19:14:21 -03:00
Christian Weichel
799e6059c4 [content-service] Implement support for S3 2022-12-09 10:29:19 -03:00
Christian Weichel
478a75e744 Switch license to AGPL 2022-12-08 13:05:19 -03:00
Pavel Tumik
75a7e4fa41 [ws-manager] improve stop workspace metric 2022-12-05 03:37:15 -03:00
Pavel Tumik
807bc84009 [ws-manager] improve start workspace metric 2022-12-05 03:37:15 -03:00
JenTing Hsiao
b0c039f7c5 ws-manager: log JSON as an object instead of a string
Signed-off-by: JenTing Hsiao <hsiaoairplane@gmail.com>
2022-11-24 10:12:53 -03:00
utam0k
d27715d4a8 ws-manager: Update golden files 2022-11-15 03:30:44 +02:00
utam0k
f613525b66 ws-manager: Allow extra time for workspace to receive SIGKILL 2022-11-15 03:30:44 +02:00
Pavel Tumik
01e79f4865 [pvc] remove fsgroup 2022-11-14 21:34:44 +02:00
Manuel Alejandro de Brito Fontes
6bb1ae2e93 Refactor probe maps to use sync.Map 2022-11-12 00:15:41 +02:00
Manuel Alejandro de Brito Fontes
6c4c785165 [ws-manager] Adjust probe timeouts 2022-11-12 00:15:41 +02:00
utam0k
f19761e2ac ws-manager: Monitor a event about deleting a pod
Signed-off-by: utam0k <k0ma@utam0k.jp>
2022-11-11 07:24:40 +02:00
Pavel Tumik
f89c8ba45c [pvc] set FSGroupChangePolicy 2022-11-10 00:39:39 +02:00
Anton Kosyakov
6533581a0e [ssh-proxy] only mark first active when user establish ssh connection
VS Code Remote SSH will reestablish a ssh connection even if window is not uesd.
It will retrigger heartbeat over and over in SSH Gateway.
SSH Gateway should only triggert heartbeat if pty sessions is present.

Co-authored-by: Pudong <tianshi8650@gmail.com>
2022-11-09 21:46:39 +02:00
Pavel Tumik
0fbb0e1f60 pass workspace type in DeleteVolumeSnapshot 2022-11-09 14:58:39 +01:00
Pavel Tumik
fb524780e4 revert delete of DeleteVolumeSnapshot fn 2022-11-09 14:58:39 +01:00
JenTing Hsiao
35ca51bd5c ws-manager: remove no used function
Signed-off-by: JenTing Hsiao <hsiaoairplane@gmail.com>
2022-11-09 14:58:39 +01:00
JenTing Hsiao
a4dacef491 ws-manager: add prebuild PVC configuration
Signed-off-by: JenTing Hsiao <hsiaoairplane@gmail.com>
2022-11-09 14:58:39 +01:00
JenTing Hsiao
d686ea9685 Add counter metrics to record error count
Add two metrics to record the `mount device failed` and `cannt mount volume`
error. So, we could know the frequency the error happened.

Signed-off-by: JenTing Hsiao <hsiaoairplane@gmail.com>
2022-11-03 10:57:07 +01:00
JenTing Hsiao
0a130f1cda Skip the cannot mount workspace event because it might be an intermediate state
Signed-off-by: JenTing Hsiao <hsiaoairplane@gmail.com>
2022-11-03 10:57:07 +01:00
Pavel Tumik
9a8151df25 refactor dispose workspace when using pvc 2022-11-03 01:16:07 +01:00
Pavel Tumik
8c36a94586 [pvc] add init container to workspace to chown pvc mount folder 2022-11-02 06:20:06 +01:00
JenTing Hsiao
6fd748c4c6 ws-manager: remove PVC once the workspace pod is unschedulable
Signed-off-by: JenTing Hsiao <hsiaoairplane@gmail.com>
2022-11-01 08:50:05 +01:00
Manuel Alejandro de Brito Fontes
4700623673 [ws-manager] Configure http client timeouts 2022-10-27 01:04:41 +02:00