The grid record includes an external_storage field which you can use to specify an external location for CA AppLogic® volumes. At present, NFS is the only supported external storage type. Also, it is not possible to modify the external storage setting of an existing grid. External storage can only be specified as part of the grid creates operation.
Accessibility of external storage is not validated as part of the creation operation by default. This is consistent with the behavior of the BFC UI, where storage can be validated with a separate operation. The API create will validate storage accessibility (and fail the create operation if the storage can't be accessed) when the checkExternalStorage flag is present in the querystring of the POST operation URI. For example:
POST /BFC/grids?checkExternalStorage
The following examples include a reference to the External Storage NFS Record.
|
Field |
Type |
Description |
|
host |
string |
The NFS server. |
|
path |
string |
The file system path exported by the NFS server. |
The following table lists the possible state values for a grid. The API State Name column lists the state as it appears in the grid_record of the API. The UI State Text column lists the descriptive text displayed in the user interface for the state value. The Description column includes more information about what is happening for the specific state.
|
API State Name |
UI State Text |
Description |
|
resource_allocated_state |
Stopped |
The BFC representation of the grid is created, but no servers are assigned to it yet. |
|
resource_getting_ready_state |
Preparing to start |
Grid is searching and reserving servers for its use. |
|
resource_getting_ready_failed_state |
Start failed |
The grid failed to acquire the number of servers specified by its minimum server hardware configuration and is finishing processing before moving to a failed state. |
|
resource_ready_state |
Stopped |
The grid has servers assigned, but is not currently running due to user choice. This state is usually reached because a user used the BFC to stop the grid. |
|
resource_wait_to_run_state |
Preparing to start |
In the case of a grid, it passes through this state very quickly and no processing was done. |
|
resource_booting_state |
Starting |
This state applies to grid creation processes and grid restart. |
|
resource_cancel_boot_state |
Canceling start, waiting for operations to complete. |
A user issued a stop or delete to the grid during start. The grid is attempting to cleanly stop this operation. |
|
resource_running_state |
Running |
The grid is up and running and the BFC has not detected any issues with how it is running. |
|
resource_waiting4_uctl_status_state |
Grid operation in progress |
This can occur for any of the following reasons: Upgrade - The BFC is in the process of upgrading the grid. Update grid - The BFC is in the process of updating the grid. Add server - In the process of reimaging a new server and adding it in to the running grid. Update power credentials - The user updated the credentials for one or more power controller. The BFC is in the process of configuring CA AppLogic® with this additional data. Apply hotfix - One or more hotfixes are being applied to the grid. Apply SSH keys - The user has performed the "Manage SSH Keys" operation and the BFC is updating the grid controller and the grid servers with the new key information. |
|
resource_uctl_breached_state |
Grid requires user action |
An attempt to perform one of the operations in resource_waiting4_uctl_status_state has failed or has reached a stopping point where the BFC needs user input. For example: Upgrade staged - Normal processing: expected when a user does a grid upgrade and selects 'Reboot Later'. Too many servers in grid - Normal processing: The number of servers in the grid exceeds the target. Waiting for a user to intervene. Not enough servers - Normal processing: The grid has fewer running servers than target. Upgrade failed - Failure: an attempt to upgrade the grid failed. Grid controller offline - CA AppLogic® not running correctly: The BFC has started the servers and handed control of the servers to CA AppLogic®. After doing so, the BFC has lost contact with the grid controller. Grid server offline - CA AppLogic® not running correctly: The grid controller is online, but one or more of the servers in the grid are not currently participating in the grid. Grid update failed - One of the various update operations has failed. Apply hotfix failed- The BFC failed in applying a hotfix to the grid. This can only happen when a grid server fails in the process of applying a hotfix or if an invalid hotfix was imported. |
|
resource_waiting4_ucth_status_state |
Remove server in progress |
The user has selected a server for removal from the grid and the grid is working on removing it. |
|
resource_ucth_breached_state |
Remove server failed |
The BFC failed to remove a specific server from the grid, or the grid currently has more servers than its Target and is waiting for a user to select which server to remove. |
|
resource_wait_to_stop_state |
Preparing to shut down |
A user has issued a "Stop" command to the grid using the BFC. The BFC is running a shutdown of the grid servers. |
|
resource_failed_running_state |
Failed, running |
The grid failed during the start process. The BFC has left the grid servers in whatever state they reached so that a person can debug the problem. |
|
resource_failed_wait_to_stop_state |
Failed - waiting to shut down |
After a stop in the resource_failed_running_state, the grid is in the process of shutting down the servers. |
|
resource_failed_ready_state |
Failed, stopped |
Grid servers are stopped. Only reached after a stop in the resource_failed_running_state. |
|
resource_failed_state |
Boot failed |
Grid failed to acquire the SLA configured minimum number of servers. |
|
resource_quarantining_wait_to_stop_state |
Quarantining - preparing to shut down |
A grid delete operation was performed on a running (or partially running) grid. The grid has gone in to the resource_quarantining_wait_for_reset_state. |
|
resource_quarantining_wait_for_reset_state |
Quarantining - preparing to clean up |
Each grid server is sanitized. The first part of each partition on the disks are zeroed out so that the partition is not accessible and the server will not boot into the grid operating system until it is reimaged. |
|
resource_quarantined_state |
Quarantined |
Grid will land here for a very short time during the delete process. The BFC infrastructure is cleaning up the last remnants of the grid in its internal data structures. |
A grid_server_configuration record contains the server hardware values for the minimum, target, and maximum number of servers to allocate for a grid plus an expression indicating the server selection criteria. There is one instance of the record for the Xen Config and another for the ESX Config.
|
Field |
Type |
Description |
|
min |
integer |
Minimum number of servers of this type required by the grid. |
|
target |
integer |
Target number of servers of this type for the grid. |
|
max |
integer |
The maximum number of servers of this type allowed in the grid. |
|
selection_criteria |
list of lists |
An expression in Conjunctive Normal Form (CNF) used to select servers for the grid based on tags that are associated with the servers. These tags are either simple tags (assigned directly to a server or servers) or smart tags (associated with servers by specifying a set of hardware requirements for the servers to be selected and assigning to the tag). In both cases, the tags specified in the selection criteria are themselves simple string tag names which are included in the CNF expression. The simplest CNF (representing a selection criteria of servers that are associated with a single tag) would be [["TagName"]]. Server selection criteria need not be specified (that is, set to null). Any server that meets the minimum requirements that are specified by the CA AppLogic® version suffices. |
|
Copyright © 2012 CA.
All rights reserved.
|
|