• No results found

Changing the NTP Server Configuration

You can change the NTP server addresses defined for HP Helion OpenStack Carrier Grade at any time after installation.

During software installation, HP Helion OpenStack Carrier Grade is configured with up to three NTP server IP addresses. You change these addresses using the web administration interface or the CLI.

Before changing NTP server addresses, review the Fault Management page and ensure that any existing system alarms are cleared.

Caution:

For the HP Helion OpenStack Carrier Grade system to use FQDN servers instead of IPv4 servers, at least one valid DNS server must be specified.

1. In the HP Helion OpenStack Carrier Grade web administration interface, open the System Configuration pane. The System Configuration pane is available from the System Panel section of the Admin menu.

2. Select the NTP tab.

The NTP page appears, showing the currently defined NTP servers and their IP addresses.

3. Click Edit NTP.

The Edit NTP dialog box appears.

4. Replace the NTP Server IP addresses or names with different ones as required.

If you specify the NTP servers using FQDNs you must have at least one valid DNS server defined on your system. You must use IP addresses otherwise.

5. Click Save.

This raises major alarms against the controllers and services. You can view the alarms on the Fault Management page.

6. Lock and unlock the controllers to clear the alarms.

Open the Hosts list by clicking Inventory on the System Panel section of the Admin menu, and then selecting the Hosts tab. Hosts requiring attention are shown with the status Config out-of-date.

To lock or unlock a host, click More for the host and then use the menu selections. a) Lock the standby controller.

Wait for the lock operation to be completed. b) Unlock the standby controller.

Wait for the host to become available. It is updated with the new NTP server addresses, and its error message is cleared.

c) Perform a swact on the active controller.

Click More > Swact Host > for the active controller. d) Lock the original controller (now in standby mode).

Wait for the lock operation to be completed. e) Unlock the original controller.

Wait for it to become available. It is updated with the new NTP server addresses, and its error message is cleared.

7. Ensure that the 250.001 Configuration out-of-date alarms are cleared for both controllers.

Changing Storage Space Allotments on the Controller

You can change the allotments for controller-based storage at any time after installation. This includes the space allotted for database, image, and backup storage, and for clusters using an LVM backend, the space allotted for Cinder volume storage.

Note:

Cinder volume storage is allotted using 3PAR, not on the controller node..

During software installation, HP Helion OpenStack Carrier Grade is configured with the following storage allotments. Database storage

The storage allotment for the OpenStack database. Image storage

The size of the partition to use for image storage. Backup storage

The storage allotment for backup operations. Volume storage

For a system using LVM storage, the storage allotment for all Cinder volumes used by guest instances. You can change these allotments using the System Configuration controls in the web administration interface, or using the CLI. For CLI instructions, see System Configuration Management Using the CLI on page 88.

To accommodate the changes, there must be enough disk space on the controller, including headroom needed to complete the operation. The headroom required is 45 GiB on the primary disk for a cluster using an LVM backend, or 65 GiB for a cluster using a 3PAR backend. This is in addition to the space required for the new allotments. The requested changes are checked against available space on the affected disks; if there is not enough, the changes are disallowed.

To provide more space, you can replace the affected disk or disks. Database, image, and backup storage use space on the primary disk. Cinder volume storage (on a cluster with an LVM backend) uses space on a disk selected by device node number during controller configuration. The replacement disk must occupy the same device node number. Changes to the Cinder volume storage may also affect the primary disk because of the headroom requirement. To pass the disk-space checks, any replacement disks must be installed before the allotments are changed.

Caution:

The configurations for controller-0 and controller-1 must be identical. If you make changes to one, you must make the same changes to the other.

Changes to the storage allotments require a reinstallation of the HP Helion OpenStack Carrier Grade host software on the controllers, even if the primary disk is not replaced.

Note:

Performing a host reinstall successfully is dependent on your BIOS boot order. Observe the following tips. • Prior to installation, the BIOS should be configured to allow booting from disk and the network.

• During the host re-installation, it is advisable to have a console serial cable attached to the host to observe the boot progress.

• BIOS boot order should be: 1. The boot partition.

Typically, this is the disk associated with /dev/sda and is as defined in /proc/cmdline when the load is booted.

2. The NIC on the boot interface (e.g. management or pxe network).

• Set BIOS boot options to ensure failsafe boot if available, e.g. rotating through available boot interfaces, watchdog timer on boot, retry boot interfaces etc.

If the BIOS boot still fails to progress, it may be necessary to force boot via the network interfaces via the BIOS boot option.

Before changing storage allotments, prepare as follows:

• Calculate your system storage requirements. Include the headroom required for changes to the storage space allotments.

• Record the current configuration settings in case they need to be restored (for example, because of an unexpected interruption during changes to the system configuration). Consult the configuration plan for your system. 1. If necessary, install replacement disks in the controllers.

If you do not need to replace disks, you can skip this step. To determine whether you need to replace disks, calculate your storage requirements. Be sure to include the headroom required on the primary disk. To replace disks in the controllers, follow these steps.

a) Lock the standby controller. b) Power down the standby controller. c) Make any required hardware changes. d) Power up the standby controller.

e) Reinstall the HP Helion OpenStack Carrier Grade software on the standby controller. f) Unlock the standby controller.

g) Swact the controllers.

h) Repeat these steps for the new standby controller. 2. Edit the disk storage allotments.

a) In the HP Helion OpenStack Carrier Grade web administration interface, open the System Configuration pane.

The System Configuration pane is available from the System Panel section of the Admin menu. b) Select the Controller Filesystem tab.

c) Click Edit Filesystem.

The Edit Controller Filesystem dialog box appears.

Note:

The Cinder storage (GiB) field is present only for clusters using the LVM backend. d) Replace the storage allotments as required.

e) Click Save.

This raises major alarms against the controllers (250.001 Configuration out-of-date). You can view the alarms on the Fault Management page. In addition, the status Config out-of-date is shown for the controllers in the Hosts list.

3. Lock the standby controller.

4. Reinstall the HP Helion OpenStack Carrier Grade software on the standby controller.

This step is required to update the system configuration, even if you have already reinstalled the HP Helion OpenStack Carrier Grade software as part of the disk replacement procedure.

5. Unlock the standby controller.

When the controller is unlocked, the 250.001 Configuration out-of-date alarms against the controller are cleared. 6. Swact the controllers.

7. Lock the new standby controller.

8. Reinstall the HP Helion OpenStack Carrier Grade software on the controller. 9. Unlock the controller.

10. Confirm that the 250.001 Configuration out-of-date alarms are cleared for both controllers.

After making these changes, ensure that the configuration plan for your system is updated with the new storage allotments and disk sizes.

System Configuration Management Using the CLI