What’s New in Nirmata 2.6
- Environmental Management Enhancements
- Cluster Resource Viewing
- Cluster Level Add-ons
- Cluster Level Prometheus
- Add-on Enhancements
- Network Policies
- Environment Panel Organization
- Cluster Panel Organization
- Alarm Panel Organization
- Jira Change Management Integration
- High Availability Cluster Enhancement
- Synchronize Unmanaged Namespaces
Fixes
- Show activity around this time' link does not display activity data’s for 5hrs and above old activities.
- Alarm clear not discernible from web hook notification
- Duplicate alarms are created sometimes
- scope information not available in web hook notification
- Add HPA form doesn’t show Stateful Sets in the Target drop-down menu
- Add support for dynamic tunnels
- After rebooting devtest2 host, nirmata-kube-controller keeps connecting/disconnecting
- Agent cannot reconnect after wss timeout
- AKS cluster is not getting created in production
- AKS Kubernetes Version is hard coded
- AKS regions are hard coded
- Alarm notification emails contain incorrect user in Last Modified by field
- Application clone does not work
- Backend allows Duplicate alarms to be added.
- Cannot delete an environment in devtest
- Cannot scale activity service
- check if cert/key directories are correct
- Check the content of cert/key, needs to be base64 encoded
- Cluster & Environments services failed to connect to ProSoft qa-cluster
- Cluster activity is showing global activity also.(in devtest2)
- ContainerVMSizeTypes is hardcoded in a class
- Creation of AKS provider managed cluster was failed.(in production account)
- Crumbs don’t render in the container template view (Uncaught TypeError)
- Deployment in “Unknown” state when pods are running
- EKS cluster created but not connected.
- Email notifications are sent for disabled alarms
- Enable alarm option is Blocked
- Endpoint label mismatch
- Environment name should allow “-”
- Environment name should not allow space.
- Environment shows Cluster Connect Status for a Disabled cluster
- Existing logic for selecting a workload controller doesn’t work properly in case of Jobs
- HA testing causes SaaS deployment to hangup when bringing back 2 nodes out of 3
- If we try to retrieve more activity data same data’s are being repeated.
- Improve events viewer
- Inconsistent arrangement of cards
- incorrect validation error for EnVars using configmap keys or secrets
- Informational alarms are emails but filter was set to Critical
- Invalid configuration is generated for EKS cluster
- Invalid DNS name is shown on the Pod
- Jira settings must be visible only in GA2 accounts, not in GA accounts
- Jobs that has been created by a Cron Job are showing as Active in the UI, even though they’re completed
- Links in the drop-down menu for Jobs and Cron Jobs (in crumbs) are not working
- Login to Nirmata intermittently fails
- nadm - add a “local-storage” command or use a dynamic hostpath provisioner
- nadm config path incorrect
- nadm scale orchestrater scaled down all services
- nadm should reserve resources for kubernetes components when deploying nirmata
- namespace discovery does works for the shopme deals deployment
- nirmata agent crash on startup
- Not able to add aws Hostgroup in devtest2.
- NullPointerException when applying ResourcePatch
- One task fails during the creation of a managed cluster on aws: failed to apply ingress controller yaml
- Pod template labels are not rendering in the Cron Job view
- Pod yaml is not loaded and the dialog just shows Fetching..
- Possible multi-tenancy issue
- Progress bar is showing only 10% while creating discovered cluster.(devtest)
- registrycreation in pending state
- Render a Workload Controller info for job pods and cron job pods in PodSummaryTabview
- Reserve resources for kubernetes containers and system services when deploying cluster
- Rest api call to fetch the AKS maximum node count associated with VM size
- Retain more logs and events (for prod environments)
- Service versions are incorrect
- signing up with google account using an already existing email doesn’t work
- Some tunnel connections were down
- State of applications takes long to update in Catalog view after adding or deleting an application (in production)
- the new buildnotify API doesn’t work
- Tunnel Client - add support for dynamic tunnels
- UI - Pods appear in the application that it doesn’t belong to
- Unable to generate API key in devtest2,getting 401 error.
- Uncaught type error in the Running Cron Job view (active crumb doesn’t show up)
- Update curator to 4.2.0 to potentially solve the problem of locks piling up.
- Use Kustomize for YAML customizations
- User is able to Delete cloud provider with connected HostGroup and Cluster
- Vault secrets should work when put in root directory
- When nirmata pods are edited, anti-affinity of required v/s preferred stops pods from running
- When running two applications in a shared namespace, failure notification of one application show in another.
- Duplicate aarm types are allowed - causing processing errors
- Failed to deploy Rook - Ceph
- addRelationList function adds an extra line that shouldn’t be added
- An asterisk sign is required for ‘Amazon Machine Image (AMI) Id’ field in the Settings-tab of Host Group.
- An asterisk sign is required for ‘Instance Type’ field in the Settings-tab of Host-Group.
- An error message displayed as “An Image is required” rather than ‘Amazon Machine Image (AMI) Id is required’ in the UserData-tab of the Host Groups.
- Application catalog should not allow application run name to have space
- Do not allow spaces in application run name
- environments service on master branch doesn’t start after changing the model
- Issue with etcd components on Dev server again
- Patch policy is not applied when a change is made in the catalog
- Pods created by Prometheus operator are not discovered
- Application catalog should not allow application run name to have space
- Issue with etcd components on Dev server again
- Deployment in “Unknown” state when pods are running
- Endpoint label mismatch
- Retain more logs and events (for prod environments)
- signing up with google account using an already existing email doesn’t work
- Vault secrets should work when put in root directory
- When nirmata pods are edited, anti-affinity of required v/s preferred stops pods from running
- Nirmata not coming up - after kubelet shutdown
- Nirmata UI having issue while bringing up controller service
- “Host not connected” on Nirmata while server looks good
- Cluster Service throw error - failing to create/update the node
- Vault Integration: Data not populated when Vault is enabled for the cluster
- kubectl commands sometimes fail
- Controller yaml issue
- 90 days of audit history
- NIBR aqua scan - fix file permissions
- Check when deleting catalog application
- Environment Services did not come up properly after bringing down 2 out of 3 master nodes in base cluster
- For a deployed application registry always points to dockerhub
- Nadm Restore Logic
- direct connect nodes in unknown status
- Vault secrets should use EmptyDir instead of hostpath
- Vault secrets should work when put in root directory
- Changes to catalog are not reflecting after push
- Issue with etcd components on Dev server again
Known Issues
- Security & webclient status servlets are throwing exception (sometimes)
- Login for SAML login is broken in 2.6.1