Bug 2089797: Cache S3 errors from UploadPV (and other MDR fixes) #44
Conversation
only when all DRPCs are deleted Signed-off-by: Jolly Mishra <jmishra@redhat.com> (cherry picked from commit ee61738)
Signed-off-by: Tesshu Flower <tflower@redhat.com> (cherry picked from commit f33fbf6)
Signed-off-by: hatfieldbrian <bhatfiel@redhat.com> (cherry picked from commit feff50f)
add to ObjectStorer interface: - UploadObject() - DownloadObject() remove from ObjectStorer interface: - UploadPV() - DownloadPVs() remove from s3ObjectStore interface: - UploadTypedObject() - VerifyPVUpload() - DownloadTypedObjects() remove: - ObjectStorer.GetName() modify fakeObjectStorer implementation: - store objects in a map - maintain one map per profile - DeleteObjects() requires specified prefix match Signed-off-by: hatfieldbrian <bhatfiel@redhat.com> (cherry picked from commit 0043954)
Initially drclusterUpdater was mainly for updating the DRCluster status. However, since the structure is used for other functionalities such as fencing/unfencing, rename the structure into an instance. Signed-off-by: Raghavendra M <raghavendra@redhat.com> (cherry picked from commit 0f84a7f)
* Initialize the status only when status cannot be found in the fetched resource from reconciler. This will avoid some reconciles happening just because of status update. * Update status only when current status is different from the previously saved status. Signed-off-by: Raghavendra M <raghavendra@redhat.com> (cherry picked from commit ae7cb92)
Signed-off-by: Raghavendra M <raghavendra@redhat.com> (cherry picked from commit 5800160)
* Update the required version of k8s.io/client-go package to 0.23.6 to maintain compatibility with csiaddons * Bump up the required go version from 1.16 to 1.17 Co-authored-by: Raghavendra Talur <raghavendra.talur@gmail.com> Signed-off-by: Raghavendra M <raghavendra@redhat.com> (cherry picked from commit d953bff)
For VRG, when the reconcile begins, status conditions are not initialized if the length of the conditions is greater than 0. This check can sometimes miss the case where one of the conditions is not there in the status, and the check assumes everything is correct. This commit changes it to check explicitly for 4 conditions for not initializing the conditions. Otherwise, it initializes a condition only if that particular condition type is not found in the status. Signed-off-by: Raghavendra M <raghavendra@redhat.com>
Currently the way kubebuilder is defined in DRCluster types, it does not consider ManuallyUnfenced option. It makes it difficult to consume the option while creating/updating the DRCluster resource. Signed-off-by: Raghavendra M <raghavendra@redhat.com>
Add tests for failover/relocation in syncDR for manual fencing and unfencing Signed-off-by: Raghavendra M <raghavendra@redhat.com>
Signed-off-by: hatfieldbrian <bhatfiel@redhat.com>
Signed-off-by: hatfieldbrian <bhatfiel@redhat.com>
- During relocate, before running a final sync - keep the check to ensure no pod is mounting the PVC - if no pod is, then do an additional check of checking for volumeattachments to make sure the underlying PV has been unmounted from nodes (and therefore I/O writes complete). If a CSI driver does not support volumeattachments then this check will not find any, and will assume the PVC is unmounted. Signed-off-by: Tesshu Flower <tflower@redhat.com>
DRCluster creates the NetworkFence resource in the managed cluster to fence/unfence a cluster. While some of the information included in the NetworkFence resource at the time of create/update is known from the Spec of the DRCluster resource (like fence state, CIDRs), there are other storage specific information that NetworkFence need to know such as storage driver, storage parameters, secrets to be able to communicate with the storage layer. That information is not part of DRCluster spec. Hence, for now, DRCluster gets that data from its annotations. In future it should be able to get that data from its status. Signed-off-by: Raghavendra M <raghavendra@redhat.com>
Initial cache of errors built in commit e7b7b9d did not account for errors faced during uploads. Emperically it is seen that connection errors are generated during the upload phase than the intial session creation phase. This commit adds updating the cache with S3 errors during the upload as well. Signed-off-by: Shyamsundar Ranganathan <srangana@redhat.com>
Signed-off-by: Shyamsundar Ranganathan <srangana@redhat.com>
* Currently MCVGetter interface can only get the MCV resource for VolumeReplicationGroup and Namespace resources. Added the functionality to get NetworkFence resource * Moved the MCVGetter code to a separate util file called mcv_util.go * Corresponding changes in DRPC to consume the MCVGetter from util * Consumption of MCV in DRCluster for NetworkFence resource Signed-off-by: Raghavendra M <raghavendra@redhat.com>
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: ShyamsundarR The full list of commands accepted by this bot can be found here.
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
|
@ShyamsundarR: This pull request references Bugzilla bug 2089797, which is valid. The bug has been moved to the POST state. The bug has been updated to refer to the pull request using the external bug tracker. 3 validation(s) were run on this bug
Requesting review from QA contact: In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
|
@openshift-ci[bot]: GitHub didn't allow me to request PR reviews from the following users: keesturam. Note that only red-hat-storage members and repo collaborators can review this PR, and authors cannot review their own PRs. In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
|
@ShyamsundarR: All pull requests linked via external trackers have merged: Bugzilla bug 2089797 has been moved to the MODIFIED state. In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
Backporting final fix for Bug 2089797 additionally following fixes for MetroDR are also included: