Skip to content
Sosivio

Release Notes

Release 1.7.1

Bug Fixes

  • License Generator Fix: Resolved an issue where the option to utilize Sosivio's license generator feature was not functioning correctly.

Release 1.7.0

New Features

  • Sosivio License Update: We are pleased to introduce the new "Core" license for Sosivio, now accessible to the general public. The "Core" license encompasses a range of advanced features, including:
    • Real-Time Metrics
    • Cluster Health Checks
    • Application Profiling
    • AI-Enabled Failure Detection
    • AI-Enabled Failure Prediction
    • AI-Enabled Failure Resolution
    • SSO (Single Sign-On), OAuth, and RBAC (Role-Based Access Control) functionality

The "Core" version is available for all clusters that fulfill the following requirements: up to 10 Nodes or up to 160 vCPU, and up to 5 users.

Release 1.6.1

Bug Fixes

  • Login Page: Fixed login issue after password change for admin users.
  • Kubernetes 1.27 Support: Updated Sosivio to be compatible with Kubernetes 1.27.

Release 1.6

New Features

Command Center page

  • Panic Indicators: Sosivio now detects application code failures, known as panics, and presents them in the command center as "Failures." This feature provides information about the location of the panic in the code. Users can easily copy the code snippet for more efficient bug reporting and resolution. Supported Languages: Sosivio supports the following programming languages for panic detection:
    • Golang
    • Python
    • NodeJS (JavaScript)
    • .Net (C#)
    • Java

0Auth Integration

  • 0Auth Login Support: Sosivio now offers login support with 0Auth without the need for RBAC (Role-Based Access Control) configuration. Users can choose whether to connect RBAC or manually create users with 0Auth login.

General

  • Cluster Reader Mode: During installation, users can now configure cluster reader permissions. Enabling this mode allows users to access all information provided by Sosivio but restricts their ability to perform any actions. For detailed information, please refer to the Sosivio documentation here.

Bug Fixes

  • Improved Sosivio microservices stability.
  • Improved Sosivio failure detection engine stability.
  • Improved Command Center failure detection.
  • Fixed login compatibility with Microsoft AzureAD for 0Auth support.

Release 1.5.3

Bug Fixes

Health Check Page

  • Health Over Time Graph: Fixed a bug where first hour scores were not displayed when viewing in the first hour scores view for a newly deployed cluster.
  • Search Bar: Fixed a bug for searching by the Health Check name.

General

  • Improved the display of Application Profiling recommendations.
  • Fixed a bug where OCP4 user with access to several namespaces could not get the namespace count.
  • Fixed the admin password being reset to the default password after an upgrade from version 1.4.2 to 1.5.2.

Release 1.5.2

General

  • Improved Sosivio data flow validation.

Release 1.5.1

Bug Fixes

Real-Time Metrics

  • Fixed Pod inner page back button: In the Pod inner page, when sorting the connections table then clicking the "back" button, sorting canceled on the connection table instead of going back to the previous page.

General

  • Known Bug: When upgrading from v1.4.2 to v1.5.1 the admin password might be reset to the one time password (OTP). To retrieve the OTP run the following command:
kubectl get secret sosivio-admin-otp -n sosivio -o jsonpath='{.data.password}' | base64 -d
  • Known Bug: The Highest Throttling Pod widget on the Insights and Metrics page might take a couple of minutes to display data.

Release 1.5

New Features

Insights & Metrics

  • Pod Page: Connection Data Table - Added connection data table providing the ability to see the Pod's connections to other Pods, services, and external entities (any entity that has an IP address) and the latency and response rate for each connection.

User Management

  • 0Auth Integration: Connect other Single Sign On (SSO) logins using OAuth and enable individual permissions for users of Sosivio within an organization. Permisssions are inherited by the users predefined roles managed whithin the users OAuth provider. For more information see here.

Bug Fixes

General

  • Known Bug: Multi Cluster View - when rejecting a cluster request the status displays as "Pending approval".
  • Known Bug: Inconsistent data between Optimization Recommendations and Application Profiling for unstable applications. An unstable application can trigger a recommendation in the Command Center, but will not appear as having a recommendation in Application Profiling page or vise versa. As a result of the unstable state of the application, submitting the recommendation might not perform any action by Sosivio.
  • Known Bug: Health Check page filtering - when trying to filter "is not" twice in the same filter there are no result displayed.
  • OpenShift 3 Support: As of Sosivio V1.5 Sosivio's installer & operator will no longer support Openshift 3.x.

Release 1.4.2

New Features

Insights & Metrics

  • Insights & Metrics: Added the Insights & Metrics page. This page provides an overview of users application’s and system’s status and the ones that need the most attention. Application Profiling and Real Time Metrics are also found in this section now.

General

  • Sosivio microservices improvements: Optimized Sosivio resource utilization and Sosivio autoscaling functionality.

Bug Fixes

Application Profiling

  • Application profiling now supports Openshift deploymentConfig objects.

General

  • Admin password during upgrades: When performing an upgrade from an earlier version of Sosivio to version 1.4.2 the admin password remains as the last used admin password and does not reset.
  • Improved Sosivio microservices stability.
  • Known Bug: Health Check page filtering - when trying to filter "is not" twice in the same filter there are no result displayed.
  • Known Bug: Multi Cluster View page - when rejecting a cluster request the status displays as "Pending approval".
  • Known Bug: SosivioDB - When installing or upgrading Sosivio it can take up to 20 minutes for Sosivio to fully adjust to the cluster size. When the final adjustment takes place a flicker of the counts of cluster resources can occur up to 5 seconds. The short flickering might occur again, but only if cluster resources increase more than 40%.
  • Known Bug: When all of SosivioDB Pods are erased or terminated a loss of users and/or license information might occur. To recover the error, run the following commands:
kubectl scale deploy actuator -n sosivio --replicas=0
kubectl scale deploy crud-manager -n sosivio --replicas=0
kubectl scale deploy sosiviodb -n sosivio --replicas=0
sleep 60
kubectl scale deploy actuator -n sosivio --replicas=1
kubectl scale deploy crud-manager -n sosivio --replicas=3
kubectl scale deploy sosiviodb -n sosivio --replicas=6

Release 1.4.1

New Features

Cluster Overview page

  • Failed Tests widget: Added a widget to display faild health checks for the Platform, Applications, and Deployment.
  • Node Load Average widget: Added a widget to display a descending list of each Node's load average.
  • Over Allocated widget: Added a widget to display a list, sorted by highest priority, of applications that have over-allocated resources.
  • Under Allocated widget: Added a widget to display a list, sorted by highest priority, of applications that have under-allocated resources.

Live Metrigs page

  • Application Profiling: Added Application Profiling page. This page provides resource allocation recommendations for the applications deployed on your Kubernetes environments.

Command Center page

  • Optimization page: Added Optimizations Recommendations page. This page allows users to implement recommended resource allocation and take action to optimize your applications behavior. In the page you will see both details of recommendations and the impact they have on your cluster.

General

  • Sosivio microservices improvement: Reduced the attack surface for operating system dependency inside the containers.

Bug Fixes

Live Metrics page

  • Fixed flickering bug: Real time metrics main table was flickering periodically.

General

  • Fixed Sosivio Microservices Resource Consumption: In some larger environments, Sosivio microservices had slightly higher resource consumption than needed.
  • Fixed Known Bug: Sosivio’s Sequence Recognition engine restarted after performing an upgrade from version 1.3.1 to version 1.4.1. This resulted in displaying some failures a "Warning", which would then disappear and be redetected after the restart was completed.
  • Fixed Admin Password Reset: After upgrading from version 1.3.1 to version 1.4.1, the admin password would occasionally be reset to the default admin password.

Release 1.3.1

Bug Fixes

General

  • Improved Sosivio microservices stability.
  • Improved Sosivio AI engine performance.

Release 1.3

New Features

General

  • Grafana dashboard support: Added a Grafana dashboard to allow users to have information from Sosivio displayed in Grafana. For more details see here.
  • Sosivio Microservice Containers: Sosivio containers are changed from CentOS containers to be scratch containers, UBI containers, and Debian containers.
  • When upgrading from v1.2 to v1.3 please allow up to 10 minutes for the Sosivio AI engines to sync. Before the sync is completed there can be missing environment information.

Live Metrics page

  • Favorites: Added a "favorites" option to the Pods, Nodes and Namespaces tabs. Mark any Pod, Node, or Namespace as a "favorite" by selecting the star on the respective line. A "favorite" Pod, Node, or Namespace will stay at the top of the list in order in which it was selected as a "favorite".
  • Node name as link: In the Pods view, Nodes are linked to the Node page. Click on the Node name in order to navigate to the Node detail inner page.

Sosivio Notifications and Alerts

  • Sosivio Alerts Configuration: Choose which types of alerts to be notified on while using Sosivio. Recieve Alerts from the Command Center, Health Check, and Multi Cluster View. Alerts will be visible in the alerts pop up widget on the bottom bar.
  • 3rd-Party Notifications: Added support for alerts to be sent to Prometheus Alertmanager, e-mail, SMS (Short Message Service),and Slack integration.

Command Center page

  • Cluster Rebalancing: When Sosivio detects excessive load on several Nodes, it will advise users to rebalance the Pods in the cluster.

Bug Fixes

General

  • Default Password: The admin user default password at is now auto generated during the inital installation. During the first login, the admin user is advised to change the auto generated password at the user management page. See details here.
  • Improved Sosivio microservices stability.

Release 1.2.1

New Features

General

  • Changes to Sosivio containers: Sosivio containers are now scratch containers.

Bug Fixes

Live Metrics page

  • Fixed bug where Node Real-time metrics graph displayed incorrect CPU metrics.
  • Fixed bug where cloud managed Kubernetes platforms displayed incorrect throttling metrics.

Command Center page

  • Fixed a bug for displaying a failure of a pod that relates to a non-exciting parent entity.

Release 1.2

New Features

General

  • Notifications pop-up: Notifications will show for all "Failures" from the command center. Clicking on the notifications will lead to the command center page. Notifications are located on the bottom bar.
  • Improved preformance of several A.I microservices which results in earlier failure detection.

Live Metrics page

  • Containers list: In Pod Data page, a list of all the Pod's containers is presented. The real-time metrics graph has a drop down list of the Pod's containers and allows you to choose which containers in the Pod to present (one, multiple, or all).
  • Pod Real-time Metrics graph: Added new live metric - Throttling. Throttling is a precentage of times tasks in a cgroup have exhausted all of their available time as specified by the CFS (Completly Fair Schedular) quota before their quota is reallocated/reset.
  • Missing limit configuration indication: Added a warning for missing limit configuration for a containers in a Pod. In the live metrics home page, you can hover over the alert sign under the memory/CPU limit column in order to see the list of containers which are missing the limit configuration.

Command Center page

  • Added the option to filter by entities.

Multi Cluster View page

  • Added the option to mark clusters as favorites. Favorite clusters will be presented at the top of the list constantly, sorted by the order they were made a favorite cluster.

Health Check page

  • Filter bar: In the health checks page, there is an added option to filter the health checks to find failed test easily.

Bug Fixes

Health Check page

  • Filtering is no longer case sensitive and now supports "like" and "not like" filtering.

Command Center page

  • Auto refresh of command center set to a 15 seconds interval before auto play is back to active.

Release 1.1.2

Bug Fixes

K8s Components page

  • Fixed OpenShift 4 issue where OVS (Open vSwitch) was inaccuratley shown as not connected.

Release 1.1.1

New Features

Command Center Page

  • Auto refresh status button: When clicking on an event's details, the auto refresh button will pause to ensure readability of the event’s details. When closing the event details, auto refresh will be active again.
  • Optimized speed and accuracy of the Sosivio failure prediction AI engine.

General

  • Improved Sosivio microservices stability.
  • Optimized Sosivio AI engine performance.
  • Added tool tips to be active at initial login.

Bug Fixes

Live Metrics Page

  • Container image name not properly displayed in Pod Live Metrics.

Multi Cluster View Page

  • Improved stability when adding a cluster in protected networks (encrypted connections).

Known Bugs

Live Metrics Page

  • PostgreSQL Container Images: Real Time Metrics for Pods, Nodes, and Namespaces might display inaccurate memory usage for Pods with PostgreSQL Container images.

    Workaround: To see the accurate memory usage for Pods with this bug, run the following command: 

kubectl exec -it <<POD_NAME>> -n <<NAMESPACE>> -- top


Release 1.0

Known Bugs

Live Metrics Page

  • Container Image Name: Real Time Metrics for Pods, Nodes, and Namespaces might display inaccurate memory usage for Pods with Container image names that contain a space. Example: myapp:version 1

    Workaround: To see the accurate memory usage for Pods with this bug, run the following command:

kubectl top pod <<POD_NAME>> -n <<NAMESPACE>> --containers

NOTE: This is fixed in Release 1.1

  • PostgreSQL Container Images: Real Time Metrics for Pods, Nodes, and Namespaces might display inaccurate memory usage for Pods with PostgreSQL Container images.

    Workaround: To see the accurate memory usage for Pods with this bug, run the following command: 

kubectl exec -it <<POD_NAME>> -n <<NAMESPACE>> -- top