@@ -10,16 +10,18 @@ Known Issues in the |k8s-op-full|
10
10
:depth: 1
11
11
:class: singlecol
12
12
13
- Disable Backups before Deleting a Resource
14
- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
13
+ Deleting a Resource with Backups Removes All Snapshots
14
+ ------------------------------------------------------
15
15
16
- When deleting a resource that has backup configured, you must disable backup first.
16
+ When you delete a resource that has backup configured, the
17
+ |k8s-op-short| terminates all backups. Deleting a resource removes all
18
+ existing snapshots without warning.
17
19
18
20
Configure Persistent Storage Correctly
19
- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
21
+ --------------------------------------
20
22
21
23
If there are no
22
- `persistent volumes <https://kubernetes.io/docs/ concepts/storage/persistent-volumes/>`__
24
+ :k8sdocs: `persistent volumes </ concepts/storage/persistent-volumes/>`
23
25
available when you create a resource, the resulting |k8s-pod| stays in
24
26
transient state and the Operator fails (after 20 retries) with the
25
27
following error:
@@ -30,15 +32,15 @@ following error:
30
32
31
33
To prevent this error, either:
32
34
33
- - Provide `persistent volumes <https://kubernetes.io/docs/concepts/storage/persistent-volumes/>`__ or
35
+ - Provide |k8s-pvs| or
34
36
- Set ``persistent : false`` for the resource
35
37
36
38
For testing only, you may also set ``persistent : false``. This
37
39
*must not be used in production*, as data is not preserved between
38
40
restarts.
39
41
40
42
Remove Resources before Removing |k8s|
41
- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
43
+ --------------------------------------
42
44
43
45
Sometimes |onprem| can diverge from |k8s|. This mostly occurs when
44
46
|k8s| resources are removed manually. |onprem| can keep displaying an
@@ -49,7 +51,7 @@ resource specification to delete resources first so no dead Automation
49
51
Agents remain.
50
52
51
53
Create Separate Namespaces for |k8s-op-short| and MongoDB Resources
52
- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
54
+ -------------------------------------------------------------------
53
55
54
56
The best strategy is to create |k8s-op-short| and its resources in
55
57
different namespaces so that the following operations would work
@@ -71,7 +73,7 @@ This would mean that it could not clean the configurations, which
71
73
would have to be done in the |application|.
72
74
73
75
Difficulties with Updates
74
- ~~~~~~~~~~~~~~~~~~~~~~~~~
76
+ -------------------------
75
77
76
78
In some cases, the |k8s-op-short| can stop receiving change events. As
77
79
this problem is hard to reproduce, the recommended workaround is to
@@ -90,14 +92,14 @@ automatically and starts working correctly:
90
92
:doc:`Kubernetes Operator installation </tutorial/install-k8s-operator>`
91
93
92
94
Machine Memory vs. Container Memory
93
- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
95
+ -----------------------------------
94
96
95
97
MongoDB versions older than 3.6.13, 4.0.9, and 4.1.9 report host system RAM, not
96
98
container RAM.
97
99
98
100
99
101
Changes to Avoid
100
- ~~~~~~~~~~~~~~~~
102
+ ----------------
101
103
102
104
The |k8s-op-short| will not be able to apply the following change on a MongoDB Deployment simultaneously:
103
105
0 commit comments