Selecting Nodes in the main menu opens the Nodes screen. The
functions on this screen will vary depending on your role:
- Viewing Virtual Node Information: The Nodes screen provides
information about the virtual nodes and allows you to restart one or
more virtual node(s). See Viewing Virtual
Node Information.
- Migrating Virtual Nodes: If you are a Tenant or Platform Administrator, then
the Nodes screen may allow you to migrate one or more
virtual node(s). See Migrating Virtual Nodes.
Viewing Virtual Node Information
The Nodes screen presents slightly different information based on the user
role, as follows:
Member View
The Nodes screen appears as shown below for Tenant Administrator and Member
users:
Note:
If you are a Tenant Administrator, then this screen may
display the
Migrate Nodes button. See
Migrating Virtual Nodes for more
information.
This screen displays the following read-only information:
- Name: Name of the node.
- Instance IP: IP address used to access that node.
- Flavor: Flavor of the node. See Virtual Node Flavors for more information about flavors and The Flavor Management Screen for
information about customizing flavors.
- Distribution: Image used to deploy the node. See The Tenant App Store Screen, The Project App Store Screen, and The App Store Screen.
- Node Role: Role that the node is playing in the cluster (such as
controller or worker).
- Cluster Name: Name of the cluster to which the virtual node belongs. This
column also displays the cluster status as an icon. The possible icons
are:
- OK: blue checkmark
- Warning: orange
exclamation point
- Error: red exclamation
point
- Rebooting: animated circles
- Reboot: Clicking the Reboot icon (arrows) reboots the selected
node. A confirmation appears. Click OK to proceed.
Platform Administrator View
The Nodes screen appears as shown below for Platform Administrator users:
This screen displays the following read-only information:
- Name: Name of the node.
- Host: Host that the virtual node is running on.
- Tenant: Tenant or project to which the virtual node belongs.
- Instance IP: IP address used to access that node.
- Flavor: Flavor of the node. See Virtual Node Flavors for more information about flavors and The Flavor Management Screen for
information about customizing flavors.
- Node Role: Role that the node is playing in the cluster (such as
controller or worker).
- Cluster Name: Name of the cluster to which the virtual node belongs. This
column also displays the cluster status as an icon. Hovering the cursor over the
icon displays a brief summary of the cluster status. The possible icons are:
- OK: blue checkmark
- Warning: orange
exclamation point
- Error: red exclamation
point
- Rebooting: animated circles
- Reboot: If available, clicking the Reboot icon (arrows) in
the Actions column reboots the selected node. A confirmation appears.
Click OK to proceed.
Migrating Virtual Nodes
The Nodes screen allows you to migrate virtual nodes/containers to different
hosts, provided that all of the following conditions are met:
- An external persistent storage resource has been configured using the
Persistent Storage tab of the Settings screen, as described in
Application Persistent Storage
Tab.
- One or more flavor(s) has been created that includes a per-node
Persistent Storage allocation. See Creating a New Flavor.
Note:
If you create a flavor that does not include a Persistent
Storage allocation and then create virtual nodes using that
flavor, then editing the flavor to include a Persistent Storage
allocation will only affect virtual nodes created after the
flavor has been edited. Virtual nodes created before adding the
Persistent Storage allocation to the flavor cannot be
migrated.
- One or more cluster(s) has been created that include one or more virtual
node(s)/container(s) that use a flavor that includes a per-node Persistent
Storage allocation. See Creating a New
Cluster.
- The deployment has enough resources on one or more alternative host(s) to
accommodate the transfer, including any placement and/or tenant constraints. See
About Tags.
If all of the above conditions are met, then the Nodes screen will display the
Migrate Nodes button.
Note:
This method of container migration allows to move
individual virtual nodes/containers from the host(s) on which
they currently reside to any other host(s) in the deployment
that have sufficient resources and that meet all placement
criteria. To migrate all of the virtual nodes from one specific
host to another host, use the
Worker Vacate function in
the
EPIC Hosts Installation screen. See
The EPIC Hosts Installation
Screen.
To migrate one or more virtual node(s):
- Verify that none of the virtual node(s) that you will be migrating are
running any jobs. A virtual node must be quiescent in order to migrate.
-
Click the Migrate Nodes button.
The Cluster Containers Migration Form popup appears.
- Use the Cluster pull-down menu to select the cluster that contains the
virtual node(s) that you want to migrate. Only virtual nodes that were created
using a flavor with a defined persistent storage allocation can be
migrated.
- Click the Nodes pull-down menu to expand the list of virtual nodes in the
selected cluster that are available for migration. To select the virtual node(s)
to migrate, you may do some or all of the following:
- Type any part of a
virtual node hostname in the field. The list of virtual nodes narrows in
real time as you type.
- Select a virtual node by clicking it. A
checkmark appears next to a selected virtual node.
- Deselect a
virtual node by clicking it. No checkmark appears for an unselected
virtual node.
- Click Select All to select all of the
eligible virtual nodes in the selected cluster.
- Click
Deselect All to deselect all of the eligible virtual nodes in
the selected cluster.
- Click Submit when you have finished selecting the virtual node(s).
The following actions will then occur:
- If the deployment has sufficient resources on one or more host(s) that meet all
tenant and cluster placement criteria (see About Tags, Tenant/Project
Tags and Constraints, and Creating a New Cluster), then:
- The Cluster Containers
Migration Form window will display the notation Cluster
container migration started successfully in a green
bar.
- The status of the cluster with the selected virtual node(s)
will change to Migrating (see Cluster Statuses) while the migration is in
progress and will then return to Ready when the migration
process has completed. The cluster may display one or more warning(s)
for several minutes; this is normal behavior.
- If there are insufficient resources and/or if an error occurs, then an
error message will appear in a red bar and the migration will fail.