feat:finish backup and restore #2968
Conversation
WalkthroughA new command for the Changes
Possibly related PRs
Suggested labels
Suggested reviewers
📜 Recent review detailsConfiguration used: CodeRabbit UI 📒 Files selected for processing (1)
🚧 Files skipped from review as they are similar to previous changes (1)
Thank you for using CodeRabbit. We offer it for free to the OSS community and would appreciate your support in helping us grow. If you find it useful, would you consider giving us a shout-out on your favorite social media? 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments. CodeRabbit Commands (Invoked using PR comments)
Other keywords and placeholders
CodeRabbit Configuration File (
|
github-actions
bot
added
Invalid PR Title
📒 Documentation
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 12
🧹 Outside diff range and nitpick comments (6)
tools/kubeblocks_helm/BackupRepo_config (1)
3-8: Improve command documentation with parameter descriptionsThe placeholder values (
<...>) should be documented with:
- Expected format and constraints for each parameter
- Whether parameters are optional or required
- Example values for common scenarios
tools/kubeblocks_helm/pika-master-slave/templates/backuppolicytemplate.yaml (2)
22-26: Review backup schedule and retention policyConsider the following for the backup schedule:
- Weekly backups might not be sufficient for production workloads
- 3-month retention period (
3m) - ensure this meets compliance requirements- Document the rationale for these choices in comments
10-11: Fix YAML formatting issuesRemove trailing spaces in line 11 as flagged by yamllint.
- - componentDefs: - - pika + - componentDefs: + - pika🧰 Tools
🪛 yamllint (1.35.1)
[error] 11-11: trailing spaces
(trailing-spaces)
tools/kubeblocks_helm/pika-master-slave/dataprotection/restore.sh (1)
6-11: Enhance directory validation error messageThe error message could be more descriptive to help users understand and resolve the issue.
Consider updating the error message:
- echo "${DATA_DIR} is not empty! Please make sure that the directory is empty before restoring the backup." + echo "Error: ${DATA_DIR} contains existing data and no .kb-data-protection file found." + echo "This could indicate:" + echo "1. An incomplete previous restore" + echo "2. Existing data that needs to be backed up first" + echo "Please backup or clear the directory before proceeding with restore."tools/kubeblocks_helm/pika-master-slave/templates/backupactionset.yaml (2)
11-14: Consider adding validation for environment variablesThe environment variables are critical for backup/restore operations but lack validation.
Consider adding a validation mechanism in the scripts to ensure these environment variables are set and contain valid values. Additionally, document the expected format and constraints for each variable in comments.
21-23: Review progress sync intervalThe 5-second progress sync interval might be too frequent for large backups.
Consider making the interval configurable through a template value:
syncProgress: enabled: true - intervalSeconds: 5 + intervalSeconds: {{ .Values.backup.syncIntervalSeconds | default 5 }}
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
📒 Files selected for processing (13)
tools/kubeblocks_helm/BackupRepo_config(1 hunks)tools/kubeblocks_helm/README.md(1 hunks)tools/kubeblocks_helm/install.sh(0 hunks)tools/kubeblocks_helm/install_ms.sh(1 hunks)tools/kubeblocks_helm/installdebug.sh(0 hunks)tools/kubeblocks_helm/pika-master-slave-cluster/templates/role.yaml(1 hunks)tools/kubeblocks_helm/pika-master-slave/dataprotection/backup.sh(1 hunks)tools/kubeblocks_helm/pika-master-slave/dataprotection/restore.sh(1 hunks)tools/kubeblocks_helm/pika-master-slave/templates/backupactionset.yaml(1 hunks)tools/kubeblocks_helm/pika-master-slave/templates/backuppolicytemplate.yaml(1 hunks)tools/kubeblocks_helm/pika-master-slave/templates/componentdefinition-pika.yaml(2 hunks)tools/kubeblocks_helm/uninstall.sh(1 hunks)tools/kubeblocks_helm/uninstall_ms.sh(1 hunks)
💤 Files with no reviewable changes (2)
- tools/kubeblocks_helm/installdebug.sh
- tools/kubeblocks_helm/install.sh
🧰 Additional context used
🪛 Shellcheck (0.10.0)
tools/kubeblocks_helm/install_ms.sh
[error] 1-1: Tips depend on target shell and yours is unknown. Add a shebang or a 'shell' directive.
(SC2148)
tools/kubeblocks_helm/pika-master-slave/dataprotection/backup.sh
[warning] 29-29: Use 'cd ... || exit' or 'cd ... || return' in case cd fails.
(SC2164)
tools/kubeblocks_helm/pika-master-slave/dataprotection/restore.sh
[error] 1-1: Tips depend on target shell and yours is unknown. Add a shebang or a 'shell' directive.
(SC2148)
tools/kubeblocks_helm/uninstall.sh
[error] 1-1: Tips depend on target shell and yours is unknown. Add a shebang or a 'shell' directive.
(SC2148)
tools/kubeblocks_helm/uninstall_ms.sh
[error] 1-1: Tips depend on target shell and yours is unknown. Add a shebang or a 'shell' directive.
(SC2148)
🪛 yamllint (1.35.1)
tools/kubeblocks_helm/pika-master-slave/templates/backuppolicytemplate.yaml
[error] 8-8: syntax error: could not find expected ':'
(syntax)
[error] 11-11: trailing spaces
(trailing-spaces)
tools/kubeblocks_helm/pika-master-slave/templates/backupactionset.yaml
[error] 8-8: syntax error: could not find expected ':'
(syntax)
tools/kubeblocks_helm/pika-master-slave/templates/componentdefinition-pika.yaml
[error] 51-51: trailing spaces
(trailing-spaces)
[error] 113-113: trailing spaces
(trailing-spaces)
🔇 Additional comments (8)
tools/kubeblocks_helm/install_ms.sh (1)
1-1: Verify chart existence and dependencies
Let's verify if the referenced Helm charts exist in the repository.
✅ Verification successful
Both Helm charts exist and have valid structure
The verification confirms that both Helm charts referenced in the install script exist and have the required structure:
./tools/kubeblocks_helm/pika-master-slave/contains Chart.yaml, values.yaml, and templates directory./tools/kubeblocks_helm/pika-master-slave-cluster/contains Chart.yaml, values.yaml, and templates directory
Both charts appear to be properly structured Helm charts with all the necessary components, making the installation command valid.
🏁 Scripts executed
The following scripts were executed for the analysis:
Script:
#!/bin/bash
# Description: Check if the Helm charts exist and have the required structure
# Check for chart directories
echo "Checking chart directories..."
fd -t d "pika-master-slave$" -x echo "Found chart directory: {}"
# Check for Chart.yaml files
echo -e "\nChecking Chart.yaml files..."
fd "Chart.yaml" -p "pika-master-slave" -x echo "Found Chart.yaml: {}"
# Check for values.yaml files
echo -e "\nChecking values.yaml files..."
fd "values.yaml" -p "pika-master-slave" -x echo "Found values.yaml: {}"
Length of output: 696
Script:
#!/bin/bash
# Let's check the content of the found directory and its parent
echo "Checking contents of the helm directory..."
ls -la ./tools/kubeblocks_helm/
echo -e "\nChecking if charts exist..."
ls -la ./tools/kubeblocks_helm/pika-master-slave/ 2>/dev/null || echo "pika-master-slave directory not accessible"
ls -la ./tools/kubeblocks_helm/pika-master-slave-cluster/ 2>/dev/null || echo "pika-master-slave-cluster directory not accessible"
# Let's also check if these might be helm chart archives (.tgz)
echo -e "\nChecking for helm chart archives..."
fd -e tgz . ./tools/kubeblocks_helm/ -x echo "Found chart archive: {}"
Length of output: 2044
🧰 Tools
🪛 Shellcheck (0.10.0)
[error] 1-1: Tips depend on target shell and yours is unknown. Add a shebang or a 'shell' directive.
(SC2148)
tools/kubeblocks_helm/uninstall.sh (1)
1-1: 🛠️ Refactor suggestion
Essential improvements needed for robust uninstallation
The current implementation needs several improvements for reliability and maintainability:
Apply these changes to make the script more robust:
+#!/bin/bash
+set -eo pipefail
+
+# This script uninstalls the Pika cluster and its components
+
+echo "Uninstalling Pika cluster components..."
+
+# Function to safely uninstall a helm release
+uninstall_release() {
+ local release=$1
+ if helm status "$release" >/dev/null 2>&1; then
+ echo "Uninstalling $release..."
+ helm uninstall "$release" || echo "Failed to uninstall $release"
+ else
+ echo "Release $release not found, skipping..."
+ fi
+}
+
+# Uninstall releases
+uninstall_release "pika-cluster"
+uninstall_release "pika"
+
+echo "Uninstallation completed"The improvements include:
- Added shebang line for proper shell specification
- Added error handling with
set -eo pipefail - Added release existence validation
- Added informative output messages
- Modularized uninstallation logic into a function
- Separated operations for better error handling
Let's verify if there are any additional cleanup steps needed:
✅ Verification successful
Let me gather more information about potential cleanup requirements by checking the Helm chart and related files.
Let me check the cluster definition and component templates to understand what resources need cleanup.
Enhanced uninstallation script with comprehensive resource cleanup
+#!/bin/bash
+set -eo pipefail
+
+# This script uninstalls Pika cluster and its components, including cleanup of associated resources
+
+echo "Starting Pika cluster uninstallation..."
+
+# Function to safely uninstall a helm release
+uninstall_release() {
+ local release=$1
+ if helm status "$release" >/dev/null 2>&1; then
+ echo "Uninstalling $release..."
+ helm uninstall "$release" || echo "Failed to uninstall $release"
+ else
+ echo "Release $release not found, skipping..."
+ fi
+}
+
+# Function to cleanup resources
+cleanup_resources() {
+ local namespace=$1
+ local component=$2
+
+ echo "Cleaning up resources for $component..."
+
+ # Clean up PVCs
+ kubectl delete pvc -l app.kubernetes.io/instance="$component" -n "$namespace" --ignore-not-found=true
+
+ # Clean up ConfigMaps
+ kubectl delete cm -l app.kubernetes.io/instance="$component" -n "$namespace" --ignore-not-found=true
+
+ # Clean up Services
+ kubectl delete svc -l app.kubernetes.io/instance="$component" -n "$namespace" --ignore-not-found=true
+}
+
+NAMESPACE=${NAMESPACE:-default}
+
+# Uninstall Helm releases
+uninstall_release "pika-cluster"
+uninstall_release "pika"
+
+# Cleanup resources
+cleanup_resources "$NAMESPACE" "pika-cluster"
+cleanup_resources "$NAMESPACE" "pika"
+
+echo "Uninstallation and cleanup completed"The improved script includes:
- Proper shell environment setup
- Comprehensive cleanup of PVCs, ConfigMaps, and Services
- Namespace-aware resource cleanup
- Error handling and status messages
- Resource cleanup for both pika and pika-cluster components
🏁 Scripts executed
The following scripts were executed for the analysis:
Script:
#!/bin/bash
# Check for any potential leftover resources
echo "Checking for potential leftover resources..."
# Check for PVCs
kubectl get pvc -l app.kubernetes.io/instance=pika-cluster 2>/dev/null || true
kubectl get pvc -l app.kubernetes.io/instance=pika 2>/dev/null || true
# Check for ConfigMaps
kubectl get cm -l app.kubernetes.io/instance=pika-cluster 2>/dev/null || true
kubectl get cm -l app.kubernetes.io/instance=pika 2>/dev/null || true
Length of output: 352
Script:
#!/bin/bash
# Check for Helm chart files and resources
echo "Checking Helm chart resources..."
# Look for helm chart files
fd -e yaml -e tpl . tools/kubeblocks_helm
# Check for any cleanup or post-delete hooks in values files
rg -A 5 "helm.sh/hook.*delete" tools/kubeblocks_helm
# Check for any CRDs or other resources that might need cleanup
rg -l "kind: CustomResourceDefinition" tools/kubeblocks_helm
Length of output: 3215
Script:
#!/bin/bash
# Check cluster and component definitions for resource types
echo "Analyzing cluster resources..."
# Check cluster template for resource types
cat tools/kubeblocks_helm/pika-cluster/templates/cluster.yaml
# Check component definitions for resource types
cat tools/kubeblocks_helm/pika/templates/componentdefinition-pika-group.yaml
cat tools/kubeblocks_helm/pika/templates/componentdefinition-codis-proxy.yaml
cat tools/kubeblocks_helm/pika/templates/componentdefinition-pika-etcd.yaml
# Look for any service or statefulset definitions
rg -l "kind: (Service|StatefulSet)" tools/kubeblocks_helm
Length of output: 15075
🧰 Tools
🪛 Shellcheck (0.10.0)
[error] 1-1: Tips depend on target shell and yours is unknown. Add a shebang or a 'shell' directive.
(SC2148)
tools/kubeblocks_helm/uninstall_ms.sh (1)
1-1: 🛠️ Refactor suggestion
Reduce code duplication with uninstall.sh
This script has identical issues to uninstall.sh and performs similar operations. Consider consolidating both scripts to reduce maintenance overhead.
Create a single unified script:
+#!/bin/bash
+set -eo pipefail
+
+# This script uninstalls Pika cluster components
+# Usage: ./uninstall.sh [--all|--master-slave]
+
+# Function to safely uninstall a helm release
+uninstall_release() {
+ local release=$1
+ if helm status "$release" >/dev/null 2>&1; then
+ echo "Uninstalling $release..."
+ helm uninstall "$release" || echo "Failed to uninstall $release"
+ else
+ echo "Release $release not found, skipping..."
+ fi
+}
+
+case "${1:-}" in
+ --all)
+ echo "Uninstalling all Pika components..."
+ uninstall_release "pika-cluster"
+ uninstall_release "pika"
+ uninstall_release "pika-master-slave-cluster"
+ uninstall_release "pika-master-slave"
+ ;;
+ --master-slave)
+ echo "Uninstalling Pika master-slave components..."
+ uninstall_release "pika-master-slave-cluster"
+ uninstall_release "pika-master-slave"
+ ;;
+ *)
+ echo "Usage: $0 [--all|--master-slave]"
+ exit 1
+ ;;
+esac
+
+echo "Uninstallation completed"Consider these architectural improvements:
- Move all Helm-related scripts to a common directory
- Create a shared library for common functions
- Add documentation about the relationship between these components and the backup/restore feature
Let's verify the integration with backup functionality:
✅ Verification successful
Let me gather more information about the codebase structure and related scripts.
Based on the gathered information, I can now provide the final response:
Consolidate uninstall scripts and add backup cleanup
The review suggestion is valid but needs enhancement. The master-slave variant has additional backup-related components that need proper cleanup as shown in pika-master-slave/templates/backupactionset.yaml and backuppolicytemplate.yaml. Here's a refined suggestion:
+#!/bin/bash
+set -eo pipefail
+
+# This script uninstalls Pika cluster components
+# Usage: ./uninstall.sh [--all|--master-slave]
+
+# Function to safely uninstall a helm release
+uninstall_release() {
+ local release=$1
+ if helm status "$release" >/dev/null 2>&1; then
+ echo "Uninstalling $release..."
+ helm uninstall "$release" || echo "Failed to uninstall $release"
+ else
+ echo "Release $release not found, skipping..."
+ fi
+}
+
+# Function to cleanup backup resources
+cleanup_backups() {
+ echo "Cleaning up backup resources..."
+ kubectl delete backups.dataprotection.kubeblocks.io --all 2>/dev/null || true
+ kubectl delete backuppolicies --all 2>/dev/null || true
+}
+
+case "${1:-}" in
+ --all)
+ echo "Uninstalling all Pika components..."
+ uninstall_release "pika-cluster"
+ uninstall_release "pika"
+ uninstall_release "pika-master-slave-cluster"
+ uninstall_release "pika-master-slave"
+ cleanup_backups
+ ;;
+ --master-slave)
+ echo "Uninstalling Pika master-slave components..."
+ uninstall_release "pika-master-slave-cluster"
+ uninstall_release "pika-master-slave"
+ cleanup_backups
+ ;;
+ *)
+ echo "Usage: $0 [--all|--master-slave]"
+ exit 1
+ ;;
+esac
+
+echo "Uninstallation completed"Key improvements:
- Added backup resource cleanup for master-slave variant
- Maintained alignment with documented uninstall procedures in README.md
- Preserved the order of uninstallation (cluster first, then definition)
🏁 Scripts executed
The following scripts were executed for the analysis:
Script:
#!/bin/bash
# Check if there are any backup resources that need cleanup
echo "Checking for backup resources..."
# Check for backup CRDs and resources
kubectl get backups.dataprotection.kubeblocks.io 2>/dev/null || true
kubectl get backuprepositories.dataprotection.kubeblocks.io 2>/dev/null || true
Length of output: 201
Script:
#!/bin/bash
# Check for related helm scripts and their structure
fd -t f . tools/kubeblocks_helm/
# Look for any backup related code
rg -l "backup" tools/kubeblocks_helm/
# Check content of uninstall.sh for comparison
cat tools/kubeblocks_helm/uninstall.sh
# Look for any documentation about these scripts
cat tools/kubeblocks_helm/README.md 2>/dev/null || echo "No README found"
Length of output: 6398
🧰 Tools
🪛 Shellcheck (0.10.0)
[error] 1-1: Tips depend on target shell and yours is unknown. Add a shebang or a 'shell' directive.
(SC2148)
tools/kubeblocks_helm/pika-master-slave-cluster/templates/role.yaml (2)
21-27: Validate pod access requirements
The role grants broad read access to all pods. Consider:
- Limiting pod access to specific labels/names
- Documenting why pod-level access is needed for backup operations
✅ Verification successful
Pod access permissions are appropriately scoped
The pod access permissions in the role are justified and appropriately scoped:
- The values.yaml explicitly documents that RBAC permissions are "used by cluster component pod" including event creation
- The role is tightly bound to a specific service account via RoleBinding in the same namespace
- The permissions are read-only (get, list, watch) without any modification rights
- The role is namespace-scoped rather than cluster-wide
🏁 Scripts executed
The following scripts were executed for the analysis:
Script:
#!/bin/bash
# Search for pod interactions in backup/restore logic
rg -l "pods" | grep -i "backup\|restore"
Length of output: 40
Script:
#!/bin/bash
# Let's first check the broader context of the role.yaml file
cat tools/kubeblocks_helm/pika-master-slave-cluster/templates/role.yaml
# Then search for any pod-related operations in the codebase
rg -A 5 -B 5 "pods.*get|pods.*list|pods.*watch"
# Also check if there are any comments or documentation about pod access
rg -l "pod.*access|access.*pod" --type md
Length of output: 1398
Script:
#!/bin/bash
# Let's check what this role is used for by examining the service account bindings
cat tools/kubeblocks_helm/pika-master-slave-cluster/templates/rolebinding.yaml
# Let's also check if there are any other related RBAC files
fd "role.*\.yaml|rolebinding.*\.yaml" tools/kubeblocks_helm/pika-master-slave-cluster/templates/
# Check for any pod-related operations in the chart's values or other template files
rg -A 3 "pod" tools/kubeblocks_helm/pika-master-slave-cluster/
Length of output: 1442
14-20: Review cluster-level permissions scope
The role grants broad read access to all clusters. Consider:
- Limiting scope to specific cluster names if possible
- Adding resource names restriction for better granularity
✅ Verification successful
Let me gather more context about the cluster-related permissions in the codebase.
Let me check the rolebinding to understand how this role is being used and its scope.
Let me check the helpers template to understand how the cluster name is scoped.
The current RBAC permissions are appropriately scoped
The role and rolebinding are already well-scoped for the following reasons:
- The role is namespace-scoped (not cluster-wide)
- The role name is templated with specific cluster name (
kb-{{ include "clustername" . }}) - The permissions are minimal and read-only for clusters (get/list/watch)
- The role is bound to a specific service account that's unique to the cluster instance
- The implementation follows the principle of least privilege by only granting necessary permissions
🏁 Scripts executed
The following scripts were executed for the analysis:
Script:
#!/bin/bash
# Check for other roles/bindings that might grant similar permissions
rg -l "apps.kubeblocks.io" | grep -i "role\|binding"
Length of output: 120
Script:
#!/bin/bash
# Check the complete role definition and other RBAC-related configurations
cat tools/kubeblocks_helm/pika-master-slave-cluster/templates/role.yaml
# Look for any resource name restrictions or selectors
rg "resourceNames:" tools/kubeblocks_helm/pika-master-slave-cluster/templates/role.yaml
# Check if there are any rolebindings to understand the scope of access
fd rolebinding -t f tools/kubeblocks_helm/pika-master-slave-cluster/templates/
Length of output: 823
Script:
#!/bin/bash
# Check the rolebinding configuration
cat tools/kubeblocks_helm/pika-master-slave-cluster/templates/rolebinding.yaml
# Check if there are any cluster-specific templates or values that might restrict scope
fd "values.yaml|_helpers.tpl" -t f tools/kubeblocks_helm/pika-master-slave-cluster/
Length of output: 745
Script:
#!/bin/bash
# Check the helpers template for cluster name scoping
cat tools/kubeblocks_helm/pika-master-slave-cluster/templates/_helpers.tpl
# Check values.yaml for any default configurations
cat tools/kubeblocks_helm/pika-master-slave-cluster/values.yaml
Length of output: 3077
tools/kubeblocks_helm/pika-master-slave/templates/backuppolicytemplate.yaml (1)
14-21: Review backup method configuration
The backup configuration has potential concerns:
snapshotVolumes: falsemeans no volume snapshots - ensure this aligns with data consistency requirements- Verify that the
/datamountPath contains all necessary data for a complete backup
✅ Verification successful
Let's search for volume configurations more broadly to understand the data persistence and backup setup in the Pika templates.
Backup configuration is correctly aligned with data persistence setup
The backup configuration is properly configured based on the following findings:
- The
/datamountPath in backup policy matches the primary data volume mount in the Pika component definition - The volume mount is consistently used across containers (init, main pika container) for data persistence
- The backup action set (pika-backup) is configured to use the same
/datadirectory and includes proper backup/restore scripts snapshotVolumes: falseis appropriate since Pika's data files are being backed up directly from the mounted volume
🏁 Scripts executed
The following scripts were executed for the analysis:
Script:
#!/bin/bash
# Search for volume mount configurations
rg -l "mountPath.*data" | grep -i "volume\|mount"
Length of output: 51
Script:
#!/bin/bash
# Search for volume mount configurations with broader pattern
rg -l "/data" --type yaml
# Search for volume configurations in pika templates
fd . "tools/kubeblocks_helm/pika-master-slave/templates" -e yaml -x cat {}
Length of output: 6294
tools/kubeblocks_helm/pika-master-slave/templates/backupactionset.yaml (1)
19-20: Verify Redis image security implications
Running backup operations with the Redis image on target pod nodes requires careful security consideration.
Let's verify the image configuration and permissions:
✅ Verification successful
Redis backup operations appear safe with minimal privileges
The backup process performs standard Redis operations (BGSAVE) and file operations within a confined data directory. The script uses basic shell commands and doesn't require elevated privileges. Key observations:
- Uses redis-cli for database operations
- File operations limited to specific directories (log, db)
- No privileged operations or system-level modifications detected
- Runs with standard container permissions
While runOnTargetPodNode is set to true, the operations performed are safe and contained within the expected scope of a backup process.
🏁 Scripts executed
The following scripts were executed for the analysis:
Script:
#!/bin/bash
# Check Redis image references in the codebase
rg -l 'image:.*redis' --type yaml
# Look for any security context configurations
rg -l 'securityContext' --type yaml
Length of output: 336
Script:
#!/bin/bash
# Check the content of files with Redis image references to understand the security context
cat tools/kubeblocks_helm/pika-master-slave/templates/backupactionset.yaml tools/kubeblocks_helm/pika-master-slave/templates/componentdefinition-pika.yaml
# Check for privilege-related configurations in the backup action
rg -A 5 'runOnTargetPodNode' --type yaml
Length of output: 4952
Script:
#!/bin/bash
# Check the backup script content to understand what operations are performed
cat tools/kubeblocks_helm/pika-master-slave/dataprotection/backup.sh
# Check if there are any specific security configurations for the backup process
rg -A 10 'securityContext|capabilities|privileged' tools/kubeblocks_helm/pika-master-slave/templates/
Length of output: 1153
tools/kubeblocks_helm/pika-master-slave/templates/componentdefinition-pika.yaml (1)
28-34: LGTM! Well-defined role configuration.
The roles are properly defined with clear attributes for master and slave nodes, following Kubernetes best practices.
| @@ -0,0 +1 @@ | |||
| helm install pika-master-slave ./pika-master-slave && helm install pika-master-slave-cluster ./pika-master-slave-cluster | |||
There was a problem hiding this comment.
🛠️ Refactor suggestion
Consider adding pre-installation checks and cleanup options
The script should verify the environment and provide cleanup options:
- Check if Helm is installed
- Check if charts are already installed
- Provide options to clean up existing installations
Would you like me to provide an enhanced implementation with these features?
🧰 Tools
🪛 Shellcheck (0.10.0)
[error] 1-1: Tips depend on target shell and yours is unknown. Add a shebang or a 'shell' directive.
(SC2148)
Add shebang and proper error handling
The script needs several improvements for reliability and portability:
- Missing shebang line
- No error handling
- Relative paths could fail depending on execution context
Consider this improved implementation:
+#!/bin/bash
+set -eo pipefail
+
+# Script directory
+SCRIPT_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
+
+# Function to check if chart exists
+check_chart() {
+ local chart_path="$1"
+ if [ ! -d "$chart_path" ]; then
+ echo "Error: Chart not found at $chart_path"
+ exit 1
+ fi
+}
+
+# Check charts exist
+check_chart "${SCRIPT_DIR}/pika-master-slave"
+check_chart "${SCRIPT_DIR}/pika-master-slave-cluster"
+
+# Install charts with error handling
+echo "Installing pika-master-slave chart..."
+if ! helm install pika-master-slave "${SCRIPT_DIR}/pika-master-slave"; then
+ echo "Failed to install pika-master-slave chart"
+ exit 1
+fi
+
+echo "Installing pika-master-slave-cluster chart..."
+if ! helm install pika-master-slave-cluster "${SCRIPT_DIR}/pika-master-slave-cluster"; then
+ echo "Failed to install pika-master-slave-cluster chart"
+ exit 1
+fi
+
+echo "Installation completed successfully"
-helm install pika-master-slave ./pika-master-slave && helm install pika-master-slave-cluster ./pika-master-slave-cluster📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| helm install pika-master-slave ./pika-master-slave && helm install pika-master-slave-cluster ./pika-master-slave-cluster | |
| #!/bin/bash | |
| set -eo pipefail | |
| # Script directory | |
| SCRIPT_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )" | |
| # Function to check if chart exists | |
| check_chart() { | |
| local chart_path="$1" | |
| if [ ! -d "$chart_path" ]; then | |
| echo "Error: Chart not found at $chart_path" | |
| exit 1 | |
| fi | |
| } | |
| # Check charts exist | |
| check_chart "${SCRIPT_DIR}/pika-master-slave" | |
| check_chart "${SCRIPT_DIR}/pika-master-slave-cluster" | |
| # Install charts with error handling | |
| echo "Installing pika-master-slave chart..." | |
| if ! helm install pika-master-slave "${SCRIPT_DIR}/pika-master-slave"; then | |
| echo "Failed to install pika-master-slave chart" | |
| exit 1 | |
| fi | |
| echo "Installing pika-master-slave-cluster chart..." | |
| if ! helm install pika-master-slave-cluster "${SCRIPT_DIR}/pika-master-slave-cluster"; then | |
| echo "Failed to install pika-master-slave-cluster chart" | |
| exit 1 | |
| fi | |
| echo "Installation completed successfully" |
🧰 Tools
🪛 Shellcheck (0.10.0)
[error] 1-1: Tips depend on target shell and yours is unknown. Add a shebang or a 'shell' directive.
(SC2148)
| kbcli backuprepo create my-repo \ | ||
| --provider s3 \ | ||
| --region <...>\ | ||
| --bucket <...> \ | ||
| --endpoint <...>\ | ||
| --access-key-id <...> \ | ||
| --secret-access-key <...> \ | ||
| --access-method <...> \ | ||
| --default \ |
There was a problem hiding this comment.
Consider security implications of exposing credentials in command
The command contains sensitive credentials that should not be exposed in plain text. Consider:
- Using environment variables or credential files
- Supporting AWS IAM roles for authentication
- Adding documentation about secure credential management
Example secure usage:
export S3_ACCESS_KEY=<access_key>
export S3_SECRET_KEY=<secret_key>
kbcli backuprepo create my-repo \
--provider s3 \
--region ${S3_REGION} \
--bucket ${S3_BUCKET} \
--endpoint ${S3_ENDPOINT} \
--access-key-id ${S3_ACCESS_KEY} \
--secret-access-key ${S3_SECRET_KEY} \
--access-method ${S3_ACCESS_METHOD} \
--default| backupFile="${DP_BACKUP_NAME}.tar.zst" | ||
| if [ "$(datasafed list ${backupFile})" == "${backupFile}" ]; then | ||
| datasafed pull -d zstd-fastest "${backupFile}" - | tar -xvf - -C ${DATA_DIR} | ||
| else | ||
| datasafed pull "${DP_BACKUP_NAME}.tar.gz" - | tar -xzvf - -C ${DATA_DIR} | ||
| fi |
There was a problem hiding this comment.
Verify backup file existence before restoration
The script should explicitly verify the backup file's existence before attempting restoration.
Add explicit validation:
backupFile="${DP_BACKUP_NAME}.tar.zst"
+if ! datasafed list | grep -q "${backupFile}\|${DP_BACKUP_NAME}.tar.gz"; then
+ echo "Error: Backup file not found in repository"
+ exit 1
+fi
+
if [ "$(datasafed list ${backupFile})" == "${backupFile}" ]; then📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| backupFile="${DP_BACKUP_NAME}.tar.zst" | |
| if [ "$(datasafed list ${backupFile})" == "${backupFile}" ]; then | |
| datasafed pull -d zstd-fastest "${backupFile}" - | tar -xvf - -C ${DATA_DIR} | |
| else | |
| datasafed pull "${DP_BACKUP_NAME}.tar.gz" - | tar -xzvf - -C ${DATA_DIR} | |
| fi | |
| backupFile="${DP_BACKUP_NAME}.tar.zst" | |
| if ! datasafed list | grep -q "${backupFile}\|${DP_BACKUP_NAME}.tar.gz"; then | |
| echo "Error: Backup file not found in repository" | |
| exit 1 | |
| fi | |
| if [ "$(datasafed list ${backupFile})" == "${backupFile}" ]; then | |
| datasafed pull -d zstd-fastest "${backupFile}" - | tar -xvf - -C ${DATA_DIR} | |
| else | |
| datasafed pull "${DP_BACKUP_NAME}.tar.gz" - | tar -xzvf - -C ${DATA_DIR} | |
| fi |
| set -e | ||
| set -o pipefail |
There was a problem hiding this comment.
Add shebang and improve error handling setup
The script is missing a shebang line and could benefit from stricter error handling.
Apply this diff to improve the script initialization:
+#!/bin/bash
+set -euo pipefail
-set -e
-set -o pipefail📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| set -e | |
| set -o pipefail | |
| #!/bin/bash | |
| set -euo pipefail |
🧰 Tools
🪛 Shellcheck (0.10.0)
[error] 1-1: Tips depend on target shell and yours is unknown. Add a shebang or a 'shell' directive.
(SC2148)
| sleep 1 | ||
| done | ||
|
|
||
| cd ${DATA_DIR} |
There was a problem hiding this comment.
Add error handling for directory change
The cd command should include error handling as suggested by shellcheck.
Add error handling for directory change:
-cd ${DATA_DIR}
+cd "${DATA_DIR}" || {
+ echo "Error: Failed to change to directory ${DATA_DIR}"
+ exit 1
+}📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| cd ${DATA_DIR} | |
| cd "${DATA_DIR}" || { | |
| echo "Error: Failed to change to directory ${DATA_DIR}" | |
| exit 1 | |
| } |
🧰 Tools
🪛 Shellcheck (0.10.0)
[warning] 29-29: Use 'cd ... || exit' or 'cd ... || return' in case cd fails.
(SC2164)
| if [ -d "log" ] || [ -d "db" ]; then | ||
| tar -cvf - ./log ./db | datasafed push -z zstd-fastest - "${DP_BACKUP_NAME}.tar.zst" | ||
| else | ||
| echo "no log db" | ||
| exit 1 | ||
| fi |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Improve error message and add size validation
The error message for missing directories is too brief, and there's no validation of directory sizes.
Enhance the validation and error handling:
+# Check directory existence and size
+MIN_BACKUP_SIZE=1024 # 1KB minimum
+
if [ -d "log" ] || [ -d "db" ]; then
+ total_size=$(du -sb ./log ./db 2>/dev/null | awk '{sum += $1} END {print sum}')
+ if [ "$total_size" -lt "$MIN_BACKUP_SIZE" ]; then
+ echo "Warning: Backup size ($total_size bytes) is unusually small"
+ fi
tar -cvf - ./log ./db | datasafed push -z zstd-fastest - "${DP_BACKUP_NAME}.tar.zst"
else
- echo "no log db"
+ echo "Error: Required directories 'log' and/or 'db' not found in ${DATA_DIR}"
+ echo "Please verify that the Redis instance is properly configured and running"
exit 1
fi📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| if [ -d "log" ] || [ -d "db" ]; then | |
| tar -cvf - ./log ./db | datasafed push -z zstd-fastest - "${DP_BACKUP_NAME}.tar.zst" | |
| else | |
| echo "no log db" | |
| exit 1 | |
| fi | |
| # Check directory existence and size | |
| MIN_BACKUP_SIZE=1024 # 1KB minimum | |
| if [ -d "log" ] || [ -d "db" ]; then | |
| total_size=$(du -sb ./log ./db 2>/dev/null | awk '{sum += $1} END {print sum}') | |
| if [ "$total_size" -lt "$MIN_BACKUP_SIZE" ]; then | |
| echo "Warning: Backup size ($total_size bytes) is unusually small" | |
| fi | |
| tar -cvf - ./log ./db | datasafed push -z zstd-fastest - "${DP_BACKUP_NAME}.tar.zst" | |
| else | |
| echo "Error: Required directories 'log' and/or 'db' not found in ${DATA_DIR}" | |
| echo "Please verify that the Redis instance is properly configured and running" | |
| exit 1 | |
| fi |
| while true; do | ||
| end_save=$(${connect_url} LASTSAVE) | ||
| if [ $end_save -ne $last_save ];then | ||
| break | ||
| fi | ||
| sleep 1 | ||
| done |
There was a problem hiding this comment.
Add timeout to BGSAVE wait loop
The while loop waiting for BGSAVE completion lacks a timeout mechanism.
Add a timeout to prevent infinite waiting:
+MAX_WAIT_SECONDS=300
+start_time=$(date +%s)
while true; do
end_save=$(${connect_url} LASTSAVE)
if [ $end_save -ne $last_save ];then
break
fi
+ current_time=$(date +%s)
+ if [ $((current_time - start_time)) -gt $MAX_WAIT_SECONDS ]; then
+ echo "Error: BGSAVE operation timed out after ${MAX_WAIT_SECONDS} seconds"
+ exit 1
+ fi
sleep 1
done📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| while true; do | |
| end_save=$(${connect_url} LASTSAVE) | |
| if [ $end_save -ne $last_save ];then | |
| break | |
| fi | |
| sleep 1 | |
| done | |
| MAX_WAIT_SECONDS=300 | |
| start_time=$(date +%s) | |
| while true; do | |
| end_save=$(${connect_url} LASTSAVE) | |
| if [ $end_save -ne $last_save ];then | |
| break | |
| fi | |
| current_time=$(date +%s) | |
| if [ $((current_time - start_time)) -gt $MAX_WAIT_SECONDS ]; then | |
| echo "Error: BGSAVE operation timed out after ${MAX_WAIT_SECONDS} seconds" | |
| exit 1 | |
| fi | |
| sleep 1 | |
| done |
| ### Back up and restore a cluster | ||
| Ensure that the default BackupRepo is defined | ||
| Fellow the kubeblock docs [kubeblocks](https://www.kubeblocks.io/docs/preview/user_docs/maintenance/backup-and-restore/backup/backup-repo) | ||
|
|
||
| create backup | ||
| ```bash | ||
| kbcli cluster backup pika-master-slave-cluster --method datafile | ||
| ``` | ||
|
|
||
| Select a backup and create a cluster. | ||
|
|
||
| ```bash | ||
| kbcli cluster restore <clusterName> --backup <backup-name> |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Documentation needs enhancement for better user guidance.
The backup and restore section requires several improvements:
- Add prerequisites and setup steps for BackupRepo
- Include the command to view existing backups:
kbcli cluster list-backups - Provide a complete example of the restore command with actual values
- Fix typo: "Fellow" should be "Follow"
- Add proper closure for the last code block
Here's the suggested improvement:
### Back up and restore a cluster
-Ensure that the default BackupRepo is defined
-Fellow the kubeblock docs [kubeblocks](https://www.kubeblocks.io/docs/preview/user_docs/maintenance/backup-and-restore/backup/backup-repo)
+#### Prerequisites
+1. Ensure that the default BackupRepo is defined:
+ ```bash
+ kbcli backuprepo create my-repo
+ ```
+2. Follow the Kubeblocks docs for detailed setup: [Backup Repository](https://www.kubeblocks.io/docs/preview/user_docs/maintenance/backup-and-restore/backup/backup-repo)
-create backup
+#### Create a backup
```bash
kbcli cluster backup pika-master-slave-cluster --method datafile-Select a backup and create a cluster.
+#### List available backups
+bash +kbcli cluster list-backups +
+#### Restore from a backup
-kbcli cluster restore <clusterName> --backup <backup-name>
+kbcli cluster restore test-pika --backup backup-default-pika-master-slave-cluster-20241203170414
<!-- This is an auto-generated comment by CodeRabbit -->
| roleProbe: | ||
| builtinHandler: custom | ||
| customHandler: | ||
| image: {{ include "redis.image" .}} | ||
| exec: | ||
| command: | ||
| - /bin/bash | ||
| - -c | ||
| - | | ||
| IFS=',' read -r -a pod_name_array <<< ${KB_CLUSTER_COMPONENT_POD_NAME_LIST} | ||
| # Ergodic name array | ||
| headless=".pika-master-slave-cluster-pika-headless" | ||
| for i in "${!pod_name_array[@]}"; do | ||
| pod_name=${pod_name_array[${i}]} | ||
| # Extracting the digital portion of the pod | ||
| pod_number=$(echo "${pod_name}" | grep -o '[0-9]*') | ||
| # If the pod number is greater than 0 | ||
| if (( pod_number > 0 )); then | ||
| pod0_svc=${pod_name_array[0]}${headless} | ||
| echo "exec:redis-cli -h ${pod_name}${headless} -p 9221 slaveof ${pod0_svc}:9221" | ||
| redis-cli -h "${pod_name}${headless}" -p 9221 slaveof "${pod0_svc}" 9221 | ||
| fi | ||
| done | ||
| preCondition: ComponentReady | ||
| - /bin/bash | ||
| - -c | ||
| - | | ||
| replication_info=$(redis-cli -h 127.0.0.1 -p 9221 info replication) | ||
| role=$(echo "${replication_info}" | awk -F':' '/^role:/ {print $2}' | tr -d '[:space:]') | ||
| if [ "$role"x = "master"x ]; then | ||
| echo -n "master" | ||
| else | ||
| echo -n "slave" | ||
| fi | ||
| periodSeconds: 1 | ||
| timeoutSeconds: 1 |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Enhance roleProbe reliability with proper error handling.
The current implementation has several potential reliability issues:
- No error handling for redis-cli command failures
- No timeout handling for redis-cli command
- Hardcoded port number should be parameterized
Suggested improvement:
customHandler:
image: {{ include "redis.image" .}}
exec:
command:
- /bin/bash
- -c
- |
- replication_info=$(redis-cli -h 127.0.0.1 -p 9221 info replication)
+ REDIS_PORT=${REDIS_PORT:-9221}
+ timeout 5s redis-cli -h 127.0.0.1 -p ${REDIS_PORT} info replication > /tmp/replication_info || {
+ echo "Failed to get replication info" >&2
+ exit 1
+ }
+ replication_info=$(cat /tmp/replication_info)
role=$(echo "${replication_info}" | awk -F':' '/^role:/ {print $2}' | tr -d '[:space:]')
+ if [ -z "$role" ]; then
+ echo "Failed to determine role" >&2
+ exit 1
+ }
if [ "$role"x = "master"x ]; then
echo -n "master"
else
echo -n "slave"
- fi
+ fi📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| roleProbe: | |
| builtinHandler: custom | |
| customHandler: | |
| image: {{ include "redis.image" .}} | |
| exec: | |
| command: | |
| - /bin/bash | |
| - -c | |
| - | | |
| IFS=',' read -r -a pod_name_array <<< ${KB_CLUSTER_COMPONENT_POD_NAME_LIST} | |
| # Ergodic name array | |
| headless=".pika-master-slave-cluster-pika-headless" | |
| for i in "${!pod_name_array[@]}"; do | |
| pod_name=${pod_name_array[${i}]} | |
| # Extracting the digital portion of the pod | |
| pod_number=$(echo "${pod_name}" | grep -o '[0-9]*') | |
| # If the pod number is greater than 0 | |
| if (( pod_number > 0 )); then | |
| pod0_svc=${pod_name_array[0]}${headless} | |
| echo "exec:redis-cli -h ${pod_name}${headless} -p 9221 slaveof ${pod0_svc}:9221" | |
| redis-cli -h "${pod_name}${headless}" -p 9221 slaveof "${pod0_svc}" 9221 | |
| fi | |
| done | |
| preCondition: ComponentReady | |
| - /bin/bash | |
| - -c | |
| - | | |
| replication_info=$(redis-cli -h 127.0.0.1 -p 9221 info replication) | |
| role=$(echo "${replication_info}" | awk -F':' '/^role:/ {print $2}' | tr -d '[:space:]') | |
| if [ "$role"x = "master"x ]; then | |
| echo -n "master" | |
| else | |
| echo -n "slave" | |
| fi | |
| periodSeconds: 1 | |
| timeoutSeconds: 1 | |
| roleProbe: | |
| builtinHandler: custom | |
| customHandler: | |
| image: {{ include "redis.image" .}} | |
| exec: | |
| command: | |
| - /bin/bash | |
| - -c | |
| - | | |
| REDIS_PORT=${REDIS_PORT:-9221} | |
| timeout 5s redis-cli -h 127.0.0.1 -p ${REDIS_PORT} info replication > /tmp/replication_info || { | |
| echo "Failed to get replication info" >&2 | |
| exit 1 | |
| } | |
| replication_info=$(cat /tmp/replication_info) | |
| role=$(echo "${replication_info}" | awk -F':' '/^role:/ {print $2}' | tr -d '[:space:]') | |
| if [ -z "$role" ]; then | |
| echo "Failed to determine role" >&2 | |
| exit 1 | |
| } | |
| if [ "$role"x = "master"x ]; then | |
| echo -n "master" | |
| else | |
| echo -n "slave" | |
| fi | |
| periodSeconds: 1 | |
| timeoutSeconds: 1 |
🧰 Tools
🪛 yamllint (1.35.1)
[error] 51-51: trailing spaces
(trailing-spaces)
| - name: redis-cli-runner | ||
| image: {{ include "redis.image" . }} | ||
| imagePullPolicy: {{ include "pika.imagePullPolicy" . }} | ||
| command: | ||
| - /bin/bash | ||
| - -c | ||
| - | | ||
| while true; do | ||
| hostname=$KB_POD_NAME | ||
| if [[ "$hostname" =~ -([0-9]+)$ ]]; then | ||
| ordinal=${BASH_REMATCH[1]} | ||
| if [[ $ordinal -ge 1 && $ordinal -le 9 ]]; then | ||
| replication_info=$(redis-cli -h 127.0.0.1 -p 9221 info replication) | ||
| role=$(echo "${replication_info}" | awk -F':' '/^role:/ {print $2}' | tr -d '[:space:]') | ||
| if [ "$role"x = "master"x ]; then | ||
| redis-cli -h 127.0.0.1 -p 9221 slaveof pika-master-slave-cluster-pika-0.pika-master-slave-cluster-pika-headless 9221 | ||
| fi | ||
| fi | ||
| fi | ||
| sleep 10 | ||
| done |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Improve redis-cli-runner container reliability and maintainability.
The current implementation has several areas for improvement:
- Infinite loop without proper termination
- Missing error handling for redis-cli commands
- Hardcoded sleep interval
- Magic numbers in ordinal check
- Limited logging for debugging
Suggested improvement:
- name: redis-cli-runner
image: {{ include "redis.image" . }}
imagePullPolicy: {{ include "pika.imagePullPolicy" . }}
+ env:
+ - name: REDIS_PORT
+ value: "9221"
+ - name: SLEEP_INTERVAL
+ value: "10"
+ - name: MAX_REPLICAS
+ value: "9"
command:
- /bin/bash
- -c
- |
+ set -eo pipefail
+
+ log() {
+ echo "$(date -u +"%Y-%m-%d %H:%M:%S") $1" >&2
+ }
+
while true; do
hostname=$KB_POD_NAME
if [[ "$hostname" =~ -([0-9]+)$ ]]; then
ordinal=${BASH_REMATCH[1]}
- if [[ $ordinal -ge 1 && $ordinal -le 9 ]]; then
+ if [[ $ordinal -ge 1 && $ordinal -le ${MAX_REPLICAS} ]]; then
+ log "Checking replication status for ordinal $ordinal"
- replication_info=$(redis-cli -h 127.0.0.1 -p 9221 info replication)
+ replication_info=$(timeout 5s redis-cli -h 127.0.0.1 -p ${REDIS_PORT} info replication) || {
+ log "Failed to get replication info"
+ sleep ${SLEEP_INTERVAL}
+ continue
+ }
role=$(echo "${replication_info}" | awk -F':' '/^role:/ {print $2}' | tr -d '[:space:]')
if [ "$role"x = "master"x ]; then
- redis-cli -h 127.0.0.1 -p 9221 slaveof pika-master-slave-cluster-pika-0.pika-master-slave-cluster-pika-headless 9221
+ log "Setting up replication for ordinal $ordinal"
+ timeout 5s redis-cli -h 127.0.0.1 -p ${REDIS_PORT} slaveof pika-master-slave-cluster-pika-0.pika-master-slave-cluster-pika-headless ${REDIS_PORT} || {
+ log "Failed to set up replication"
+ }
fi
fi
fi
- sleep 10
+ sleep ${SLEEP_INTERVAL}
done📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| - name: redis-cli-runner | |
| image: {{ include "redis.image" . }} | |
| imagePullPolicy: {{ include "pika.imagePullPolicy" . }} | |
| command: | |
| - /bin/bash | |
| - -c | |
| - | | |
| while true; do | |
| hostname=$KB_POD_NAME | |
| if [[ "$hostname" =~ -([0-9]+)$ ]]; then | |
| ordinal=${BASH_REMATCH[1]} | |
| if [[ $ordinal -ge 1 && $ordinal -le 9 ]]; then | |
| replication_info=$(redis-cli -h 127.0.0.1 -p 9221 info replication) | |
| role=$(echo "${replication_info}" | awk -F':' '/^role:/ {print $2}' | tr -d '[:space:]') | |
| if [ "$role"x = "master"x ]; then | |
| redis-cli -h 127.0.0.1 -p 9221 slaveof pika-master-slave-cluster-pika-0.pika-master-slave-cluster-pika-headless 9221 | |
| fi | |
| fi | |
| fi | |
| sleep 10 | |
| done | |
| - name: redis-cli-runner | |
| image: {{ include "redis.image" . }} | |
| imagePullPolicy: {{ include "pika.imagePullPolicy" . }} | |
| env: | |
| - name: REDIS_PORT | |
| value: "9221" | |
| - name: SLEEP_INTERVAL | |
| value: "10" | |
| - name: MAX_REPLICAS | |
| value: "9" | |
| command: | |
| - /bin/bash | |
| - -c | |
| - | | |
| set -eo pipefail | |
| log() { | |
| echo "$(date -u +"%Y-%m-%d %H:%M:%S") $1" >&2 | |
| } | |
| while true; do | |
| hostname=$KB_POD_NAME | |
| if [[ "$hostname" =~ -([0-9]+)$ ]]; then | |
| ordinal=${BASH_REMATCH[1]} | |
| if [[ $ordinal -ge 1 && $ordinal -le ${MAX_REPLICAS} ]]; then | |
| log "Checking replication status for ordinal $ordinal" | |
| replication_info=$(timeout 5s redis-cli -h 127.0.0.1 -p ${REDIS_PORT} info replication) || { | |
| log "Failed to get replication info" | |
| sleep ${SLEEP_INTERVAL} | |
| continue | |
| } | |
| role=$(echo "${replication_info}" | awk -F':' '/^role:/ {print $2}' | tr -d '[:space:]') | |
| if [ "$role"x = "master"x ]; then | |
| log "Setting up replication for ordinal $ordinal" | |
| timeout 5s redis-cli -h 127.0.0.1 -p ${REDIS_PORT} slaveof pika-master-slave-cluster-pika-0.pika-master-slave-cluster-pika-headless ${REDIS_PORT} || { | |
| log "Failed to set up replication" | |
| } | |
| fi | |
| fi | |
| fi | |
| sleep ${SLEEP_INTERVAL} | |
| done |
chejinge
added
3.5.6
4.0.2
and removed
📒 Documentation
#2970
Provide BackupRepo
Fellow the kubeblock docs kubeblocks
Fellow [tools/kubeblocks_helm/BackupRepo_config]
Create a Backup
View Backups
Restore a Cluster from a Backup
Select a backup and create a cluster:
View the Backup Policy of a Cluster
Summary by CodeRabbit
Summary by CodeRabbit
Release Notes
New Features
kbcli.pika-master-slaveandpika-master-slave-cluster.Documentation
Bug Fixes
Chores