Nodes
The Nodes section provides information about all the containers on the platform. It allows monitoring all nodes at once regardless of the host or environment, and sort results by numerous metrics, including CPU and RAM consumption.
You can hover over the columns and expand a drop-down menu to choose which data should be displayed. The complete list is reviewed below:
- Status - a color-labeled indicator of the node state (green - running, blue - sleeping, red - stopped) or spinner for the containers that are currently creating
- Node ID - unique identifier of a node within the user dashboard
- Name - icon, name, and version of the software stack template used during the node creation
- CTID - unique identifier of a node within the infrastructure
- Host Node - hardware server (host) where the node is physically located
- Cloudlets - scaling limit of a node (in cloudlets)
- CPU - average CPU power consumption
- Memory - average RAM consumption
- Disk - average disk space consumption
- IOPS - average IOPS
- Network - four columns for total incoming/outgoing internal/external traffic
- User ID - unique identifier of a user within the platform (click to locate automatically at the appropriate Users tab)
- Environment - name of environment this node belongs to (click to locate automatically at the appropriate Environments tab)
- vType - node’s virtualization type (CT - container, VM - virtual machine)
- Node Creation Date - date and time of the node creation
All the displayed consumption values are calculated as an average for the Interval selected in the tools panel. Except for the Network ones, which show a total consumption during the set period.
For monitoring convenience, the table can be sorted by any of these columns (e.g., by CPU or RAM) to easily detect the most loaded containers.
Node Management
In order to locate the required node, the following options from the tools panel can be used:
- choose the required hosts by ticking them in the hosts' filter drop-down list
- Search by the name of the stack/environment or the user/node ID
- select Interval to show nodes consumption during the specific period (last 5 min, 15 min, half an hour, hour)Note: The total statistics storing period on the platform is defined via the appropriate jelastic.statistic.nodes.cleanPeriod system setting in the admin panel (one hour by default).
Use the Search button to filter the list by the configured parameters and Clear to reset these filters. Enable Auto Refresh to update the displayed data every 20 seconds.
The following actions are available for the selected node:
- Migrate - migrates container(s) between hosts of the same host group (disabled if nodes from different hosts are chosen)
Note: Although the benefits of live (online) migration are evident, keep in mind that it is not suitable for all cases. We strongly recommend that you avoid using containers live migration for:
- environments/containers under high load - unexpected downtime with the “502 - environment stopped” error (usually brief, under 10 seconds)
- active database containers, Big Data - possible corruption or loss of the currently processed data due to the nature of online migration and freezes of network connections/disk IO related operations during the migration process
In order to change a host group, the whole environment should be migrated via the appropriate admin panel section.
- Statistics - displays a pop-up window with information on the node’s Memory, Processor, Network, Disk consumption
- Web SSH - accesses a node via SSH as a root user in a pop-up window (expand and choose the Open in Browser option to establish a connection in a new browser tab)
- Open - opens the currently selected node in a new browser tab
If needed, you can click on the node’s User ID or Environment links for additional management options at the appropriate admin panel sections.
Node Information
Upon selecting a particular node, an additional tab with detailed information about this container is automatically opened.
Here, you can find all the required data on the node, which may be required for management or troubleshooting.
