ORD1 CR5/CR8 data center consolidation FAQ
The ORD1 data center has been operational since 2010 and offers the full suite of Rackspace services. We are continuously working on improving efficiencies, and we regularly examine the performance and capacity of our data centers and the computer rooms within those data centers. To maximize efficiency and resource usage in ORD1, we need to consolidate zones CR5 through CR8, which are in the oldest area of the data center. Devices in the affected area will be physically moved into another area of the data center, and the legacy rooms will be closed.
Our dedicated migration team will carry out this move. The team has had widespread success in data center migrations over the past decade and continues to improve and automate its processes, ensuring a smooth migration for our customers.
Will service be interrupted?
Yes. Rackspace will schedule a six-hour or eight-hour maintenance window starting at 23:00 to 05:00 or 7:00 the following morning. We plan to return customers to service as quickly as possible within that window. During this time, all the affected devices will be offline while they are moved from their current physical location to the new location.
Are all of the devices in my environment being moved?
Because this is a consolidation and not a closure, it could be anywhere from one device to a customer’s entire environment in ORD1. It is important to note that other devices within an environment that are not being physically moved could also be impacted by this maintenance, depending on the device type that is being physically moved. For example, if we take your firewall offline, any devices logically located behind that firewall will be inaccessible while the firewall is offline. We have provided your account teams with a list of devices that could be impacted by this move. Reach out to your Account Manager for further information.
Will there be any IP address or Domain Name Service (DNS) changes?
For most customers, who stay in their allocated move group, the public IP addresses of the devices are retained and moved with the devices to the new location. In this case, you do not need to make any changes to your DNS.
Rarely, because of an environment’s configuration, we might need to change the IP address of a firewall or edge device. If a customer needs this modification, the Migration Project Manager or your Rackspace Account Manager reaches out to discuss the changes.
Exception - Rackspace Managed Backup service
The IP addresses used for the network that runs the Rackspace Managed Backup service (BackupNet/ServiceNet) change for all customers. Because Rackspace does not have access to Managed Colocation servers, Managed Colocation customers need to make the Managed Backup service IP address change themselves.
Rackspace takes responsibility for updating the Managed Backup service IP addresses for all other customers who use this service.
If you are using BackupNet/ServiceNet for any purpose other than Rackspace Managed Backup or for ObjectRocket, you need to make the necessary arrangements to move to ExNet before your move night. In ORD1, BackupNet/ServiceNet is available for use only for Rackspace Managed Backup and for ObjectRocket.
Exception - Colocation Account devices and VLAN tags:
Depending on the customer’s environment aggregation or data center locations, some customers might undergo a VLAN ID change during the migration. In most cases, this change does not have any discernible impact on customers because Rackspace manages these changes. However, Managed colocation accounts or colocation devices might have instances that are partially managed by the customer, and the customer needs to make changes within their environment. If your environment falls into one of these brackets, the migration project manager reaches out to the customer with the needed information.
What are the high-level stages of the migration?
The migration will happen in the following stages:
Leading up to your migration date, Rackspace will work with you and your account team to establish any specific requirements for the move. For example, let us know if you have a specific order in which your devices should be turned off or brought back online.
Rackspace will run non-disruptive audits on your devices before the migration date to ensure that we have the necessary access to your devices to run our shutdown scripts and identify issues that could impede a speedy return to service.
Before the migration starts, all devices are placed under alert suppression.
At 23:01, all affected devices are powered off by using an automated script. Any servers that need to be turned off in a specific order are powered down before the automated shutdown script runs between 22:00 and 23:01.
After all devices are powered off, they are removed from the old cabinets, racked in their new cabinets, and powered on.
Your configuration is tested and verified at an infrastructure level. If you have provided clear instructions in advance of the move date, further testing can be carried out by the migration team.
My solution is sensitive and must be handled carefully during a shutdown. Can you help?
Yes, we can work with you to understand what needs to be done to ensure a smooth migration, and we can help in various ways, including turning off and turning on your devices in a certain order.
I want to shut down the environment myself. Is that possible?
Yes, you may turn off your devices. However, we do not recommend this. Shutting down your own devices interrupts our Automation and ability to make the necessary back-end network changes. We ask that you inform us in advance that you plan to do this so that we can make the necessary changes manually before you shut your devices down.
I am a Managed Colocation customer, and you do not have access to my servers. How will the migration work for me?
Because Rackspace cannot access devices under the Managed Colocation terms, you must turn off your devices before the start of the maintenance window. If devices are not turned off by the time we need to remove them from the cabinets, we will remove their power cables. If you do not want Rackspace to turn on your devices when they are racked in their new cabinets, you must inform us at least seven days before your move night.
My data is replicated to another location. How will the migration affect this replication?
It depends on which devices are being moved, and the replication method that you have set up. We recommend that you tell us about any replication or log shipping so that we can work with you to ensure that it is re-established following the maintenance. You can provide us with explicit instructions, such as what to check for and what to do if it is not working as expected.
I have colocation devices that Rackspace does not have access to. How will these be migrated?
Because we do not have access to your colocation devices, you are responsible for shutting these devices down. If you contact us, Rackspace might be able to assist you, but this will be managed on a case-by-case basis. Our standard approach is to remove the power cables from the devices in their current location and reconnect them in their new location.
Who will test my environment following the maintenance?
Rackspace will test for basic connectivity. If you want us to run further tests, such as to check whether a website is online, you can provide us with the steps to follow to run these checks and with instructions about what to do if we encounter issues that we cannot resolve.
How do I prepare for the migration?
Rackspace will use all of its experience with data center migration projects to help support you during this consolidation exercise. Our dedicated migration team will run audits on your servers before the migration to identify and resolve any potential migration issues in advance. They will work with you and your account team to determine and agree to any special pre-migration and post-migration requirements or instructions that you might have for turning off and turning on your solution. The migration team will then focus on successfully executing your instructions on the night of the migration.
We strongly encourage customers to back up their data, either by using the Rackspace Managed Backup service or by making their own offsite backups. If your solution is not currently being backed up, you can talk to your account team about putting a backup solution in place before the migration date.
In addition to the backups, consider taking the following steps before the move night:
- Correct any known disk issues before migration night, and ensure there are no
disk corruption issues by running a check disk (
chkdsk). Doing so prevents any forced
chkdskoperations when your server is turned on.
- Fully install any pending Microsoft® Windows® updates before the scheduled migration. If updates stop us from being able to turn off your servers gracefully, the downtime for the entire move group could be extended.
- Perform a test reboot of your servers before the migration to ensure that applications and services start correctly. You can schedule reboots from the MyRackspace Portal.
I want to run a full backup before my migration. Can this be arranged?
Because it may not complete in time, we do not recommend that you initiate a full backup just before the migration. Approximately two hours before the migration is scheduled to start, our automated steps will stop any backup still in progress so that we can ensure a clean shutdown. If your backups have been completing successfully in the weeks before the migration, you should not need to run an additional backup. However, if you are concerned, you may raise a ticket with the Managed Backup team to review the possibility of running a differential (or incremental) backup before the migration. This task must complete before backups are stopped at 21:00.
How will you keep me informed about the migration?
You will receive a notification ticket 60 days from your move date and then receive notifications 45 and 14 days before your move date, with the final notification sent a day before. The final notification ticket updates you throughout the move night. We will issue updates to confirm that the migration has been given a “go” decision, to confirm that devices have been shut down, and to confirm that the devices are back online.
What if I need to make changes to my environment between now and the move night?
We will impose a change freeze on your environment two weeks before the move date. The change freeze applies to networking, physical hardware, disk expansions, and so on. It does not apply to data or application changes. If you are unsure whether a change that you need to make is impacted by the change freeze, contact your account team for validation or update one of the notification tickets with details about the change.
If there is an emergency during the change freeze and you cannot wait until after the migration to make a change, ensure that the project team is made aware of the change by updating one of the notification tickets. Failure to do so might result in unexpected behavior during the maintenance and could increase downtime.
Is the move date flexible? Can I choose the date of the migration?
No. Migration move dates are grouped together due to cabinet and IP address locations.
Whom should I contact if I have more questions about the migration?
Contact us in any of the following ways:
- Create a ticket in the MyRackspace Portal. Request in the ticket that it be assigned to the ORD1 Migration ticket queue.
- Ask a member of your account team for more information.
- Respond to one of the ticket notifications that you will receive before your move night.
©2020 Rackspace US, Inc.
Except where otherwise noted, content on this site is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License