|
5 | 5 | // * virt/monitoring/virt-runbooks.adoc
|
6 | 6 |
|
7 | 7 | :_mod-docs-content-type: REFERENCE
|
8 |
| -[id="virt-runbook-VirtualMachineCRCErrors"] |
9 |
| -= VirtualMachineCRCErrors |
| 8 | +[id="virt-runbook-VMStorageClassWarning"] |
| 9 | += VMStorageClassWarning |
10 | 10 |
|
11 | 11 | [discrete]
|
12 |
| -[id="meaning-virtualmachinecrcerrors"] |
| 12 | +[id="meaning-vmstorageclasswarning"] |
13 | 13 | == Meaning
|
14 | 14 |
|
15 | 15 | This alert fires when the storage class is incorrectly configured.
|
16 | 16 | A system-wide, shared dummy page causes CRC errors when data is
|
17 | 17 | written and read across different processes or threads.
|
18 | 18 |
|
19 | 19 | [discrete]
|
20 |
| -[id="impact-virtualmachinecrcerrors"] |
| 20 | +[id="impact-vmstorageclasswarning"] |
21 | 21 | == Impact
|
22 | 22 |
|
23 | 23 | A large number of CRC errors might cause the cluster to display
|
24 | 24 | severe performance degradation.
|
25 | 25 |
|
26 | 26 | [discrete]
|
27 |
| -[id="diagnosis-virtualmachinecrcerrors"] |
| 27 | +[id="diagnosis-vmstorageclasswarning"] |
28 | 28 | == Diagnosis
|
29 | 29 |
|
30 | 30 | . Navigate to *Observe* -> *Metrics* in the web console.
|
@@ -54,34 +54,10 @@ $ oc get pv <pv_name> -o=jsonpath='{.spec.storageClassName}'
|
54 | 54 | ----
|
55 | 55 |
|
56 | 56 | [discrete]
|
57 |
| -[id="mitigation-virtualmachinecrcerrors"] |
| 57 | +[id="mitigation-vmstorageclasswarning"] |
58 | 58 | == Mitigation
|
59 | 59 |
|
60 |
| -Add the `krbd:rxbounce` map option to the storage class configuration to use |
61 |
| -a bounce buffer when receiving data: |
62 |
| - |
63 |
| -[source,yaml] |
64 |
| ----- |
65 |
| -apiVersion: storage.k8s.io/v1 |
66 |
| -kind: StorageClass |
67 |
| -metadata: |
68 |
| - name: vm-sc |
69 |
| -parameters: |
70 |
| - # ... |
71 |
| - mounter: rbd |
72 |
| - mapOptions: "krbd:rxbounce" |
73 |
| -provisioner: openshift-storage.rbd.csi.ceph.com |
74 |
| -# ... |
75 |
| ----- |
76 |
| - |
77 |
| -The `krbd:rxbounce` option creates a bounce buffer to receive data. The default |
78 |
| -behavior is for the destination buffer to receive data directly. A bounce buffer |
79 |
| -is required if the stability of the destination buffer cannot be guaranteed. |
80 |
| - |
81 |
| -ifndef::openshift-rosa,openshift-dedicated[] |
82 |
| -See link:https://access.redhat.com/articles/6978371[Optimizing ODF PersistentVolumes for Windows VMs] |
83 |
| -for details. |
84 |
| -endif::openshift-rosa,openshift-dedicated[] |
| 60 | +Create a default OpenShift Virtualization storage class with the `krbd:rxbounce` map option. See link:https://access.redhat.com/articles/6978371[Optimizing ODF PersistentVolumes for Windows VMs] for details. |
85 | 61 |
|
86 | 62 | If you cannot resolve the issue, log in to the
|
87 | 63 | link:https://access.redhat.com[Customer Portal] and open a support case,
|
|
0 commit comments