Succeeding in the cloud

Creating a Cloud Servers configuration and equipping it with appropriate security, backup, application software, monitoring, and scaling can be a one-time experience, if you like what you build the first time and want to continue to use it.

Even if you don’t need to change your configuration, you should expect to interact with Rackspace occasionally, probably by logging in to your Cloud Control Panel to maintain your contact information, to assign specific roles to specific users, to see or change your billing options, and to explore new services and features.

To learn about some of the basic activities that you might need to perform in the Cloud Control Panel, begin with the Cloud Servers introduction page.

In the Rackspace cloud, you can quickly try something, decide that it isn’t quite right for you, and then try something else. It’s a flexible, reliable environment, but there are some limits: some practices are more productive than others, trouble can happen, and you may sometimes need help.

Managing your Rackspace account

When you log in to your Cloud Control Panel, you can examine and change your managed cloud configuration. For example, you can create a cloud server or delete a block storage volume. You can also examine and change your account itself in several areas:

  • Billing and Payments gives you access to your current balance, billing history, billing address, and credit card information.
  • Usage Overview shows you today’s usage-based billing for the current billing cycle, service by service. For Cloud Servers, you can see separate subtotals for hourly usage, image storage, and outgoing bandwidth.
  • Account Settings lets you change the account password, establish a security question, enable two-factor authentication, reset your API key, and maintain your contact information.
  • User Management lets you define users of your account and give them specific access, so that some can have full access to the account and some can have per-product limitations.
  • Resource Limits compares current activity to the limits established for your account and provides a method of requesting an increase of those limits.
  • Support Tickets shows all the open and closed tickets associated with your account and provides a method of creating a new ticket.
  • Upgrade Service Level identifies the support service level currently in effect for your account, compares it to other available service levels, and provides a method of requesting more information prior to making a change.

Managing role-based access to services

Using role-based access control (RBAC), you can divide responsibilities among members of your team. For example, you can enable a database administrator to schedule database backups and enable a network administrator to expand a load balancing group. The roles that make sense for your team are likely to change as your workload grows, your team grows, and you add more services to your configuration. You can see suggested role configurations at Managing: Role-Based Access Control (RBAC).

Managing expenses by limiting workload

Enforcing limits on some activities makes it easier to control costs. Because you pay for only what you use in the Rackspace Managed Cloud, limits mean that you will not be surprised by a sudden spike in billable activity. Limits can also protect you from workloads beyond the capacity of your configuration.

Limits on your account’s activity are initially set by Rackspace. You can change some limits yourself to suit your workload; to change other limits, contact Rackspace.

Absolute limits

Absolute limits control the total number of service resources that the user can possess simultaneously. For example, an absolute limit controls the amount of RAM that can be assigned in a Cloud Servers configuration.

Default absolute limits are set to provide you with a reasonable number of resources for testing and moderately-sized environments and applications. If you need additional resource capacity, you can request an increase of your absolute limits by opening a ticket in the Cloud Control Panel or by contacting Rackspace Support.

When you are logged in to the Cloud Control Panel, you can see absolute limits defined for servers, along with your current consumption of the limited resources, by clicking the Account: your Username menu and selecting Resource Limits.

Rate limits

Rate limits control the frequency at which the user can issue specific requests. Rate limits are reset after a certain amount of time passes. For example, a rate limit controls the number of GET requests that can be processed during a one-minute period.

Default rate limits are set such that most users won’t encounter them, however if needed, you can raise your rate limits by opening a ticket in the Cloud Control Panel or by contacting Rackspace Support.

Limits for specific services

The following table shows some examples of limits, with links to API documentation for more examples and complete details.

ServiceAbsolute limit exampleRate limit example
Cloud Block Storage10 TB in configNo limits
Cloud Servers100 servers in config100 POSTs per minute
Cloud ImagesNo limitsNo limits
Cloud NetworksNo limitsNo limits

The services mentioned here are not the only ones with absolute limits or rate limits. Some services have no limits. We publish the limits relevant to any service in that service’s API documentation. You can find those details by going to Rackspace Cloud Technical Documentation and searching for limits.

You can also avoid surprises in your usage-based billing with the help of several tools:

  • In the Cloud Control Panel, check current usage frequently.
  • Combine Cloud Monitoring and Cloud Intelligence to help you recognize extreme usage peaks.
  • Use Auto Scale to increase resources only when needed.

Best practices in the cloud

As you learn about and experiment with the many options available to you in the Rackspace Cloud, remember these best practices:

  • Assume failure can happen and design accordingly. Don’t rely on individual resources such as VMs and storage to always be available.
  • Decompose your application into smaller individually scalable parts. Don’t have a monolithic server that scales up as your demand grows.
  • When you must stop or reboot an active server, always do so via the operating system. Using the Cloud Control Panel to initialize a running server can lead to data loss and should generally be avoided.
  • Deploy a load balancer to allow you to more easily scale and update your application. Follow the configuration guidelines at Configure a Load Balancer.
  • Deploy backup and/or monitoring for your most important servers. Configure Cloud Backup as discussed at Rackspace Cloud Backup overview.

Other best practices relate to specific services, options, and configurations:

Best practices for orchestration

  • Always configure services and applications to restart upon reboot.
  • Use config drive and cloud-init to bootstrap your servers as described at Using Cloud-init with Rackspace cloud.
  • For Cloud Images, use the UUID where possible because the UUID of an image cannot change, whereas the name cannot be guaranteed to be constant.
  • Use base images rather than snapshots for build-time performance.
  • Automate configuration tasks whenever possible.

Best practices for security

  • Sending a high volume of email from your cloud server can downgrade its IP reputation. Use Mailgun instead, as described at Mailgun Quickstart.
  • Use SSH keys rather than passwords for Linux. Follow the procedure at Basic Cloud Server Security.

Best practices for storage

Best practices for networking

  • Use ServiceNet to communicate between Cloud Servers and Cloud Files and Cloud Databases. Use Cloud Networks for server-to-server communication.
  • For Cloud Networks, use RFC 1918, less the two ServiceNet blocks.

Troubleshooting in the cloud

Many common-sense practices that help you recognize, minimize, and prevent trouble in other situations apply to the Rackspace cloud as well:

  • When you encounter a problem, investigate it by eliminating as many unknowns as possible. Try to get the simplest possible example of what you want to achieve working first. Then introduce new features gradually, testing after each change, until you know exactly at which point the problem begins.
  • Maintain awareness of what is normal and abnormal within your configuration and within Rackspace overall. Monitor your systems and maintain historical logs so you know when key resources are not operating as usual.

Rackspace provides tools to help you automate detection of and response to unusual situations such as workload spikes:

Rackspace provides a real-time system status view that may help you understand whether a problem is unique to you or is caused by something more general.

You can also subscribe to our @Rackstatus Twitter feed for updates on maintenance or other service-impacting events.

As a managed cloud company, we’re always available for you to contact us directly. If you need help, you can reach out to us through the “Support” link at the top of the Rackspace Cloud Control Panel.

See also.

Although you can generalize basic troubleshooting practices from other kinds of computing work, some areas specific to working in the cloud may be new to you. Here are some tips which may help you understand and address cloud-related issues:

Troubleshooting connectivity

Connectivity issues can relate to client-to-server or server-to-server connections. They can also relate to connections among cloud services.

Symptom: An attempt to connect to a server fails.

Test 1: Try reaching the same server from different network locations, devices, and directions.

Diagnosis 1: If any method succeeds in connecting to the server, the problem is not in the server itself but in some aspect of the network.

Test 2: Try to access the server by its IP address (for example, instead of its DNS entry (for example,

Diagnosis 2: If the server is accessible by its IP address but not by its DNS entry, the problem relates to the DNS rather than to the server itself.

Test 3: Try to confirm basic TCP/IP connectivity by using telnet.

Diagnosis 3: If you cannot telnet to a target IP through a target port, a firewall may be blocking your access.

Symptom: An attempt to connect between Cloud Servers and another cloud service fails.

Test and prevention: Follow the guidance at Validating SSH configuration to determine how your SSH port should be configured to enable cloud services to interact.

See also

To learn more about troubleshooting or preventing connectivity issues, read these articles:

Troubleshooting server builds

You can build a cloud server much more quickly than you can order a dedicated server and have it installed and configured. The build process for cloud servers is rapid but is not instantaneous. Builds can be delayed by a variety of factors.

Symptom: A cloud server build takes an unusually long time to complete.

Test: Identify factors under your control that can explain why this cloud server build is slower than others you have performed:

  • Windows servers take longer to build than servers running other operating systems.
  • OnMetal servers take longer to build than virtual servers.
  • Servers built with software stacks take longer to build than bare servers (where a software stack may be installed later).
  • Servers built with backup enabled take longer to build than bare servers (where a backup capability may be established later).
  • Servers built from customer-saved images take longer to build than servers built from images provided by Rackspace.

Diagnosis: If any of these known causes of slower builds are true of the server that you are attempting to build, wait at least thirty minutes before rechecking for success or failure. Although build times vary, all server builds eventually either succeed or fail.

  • If a slow server build eventually succeeds, use the new server normally. A slow build does not predict any operational problems.
  • If a slow server build eventually fails, investigate the failure just as you would if it had failed quickly.

Symptom: A cloud server build fails.

Test 1: Check for any indication that you are being affected by a general problem.

Diagnosis 1: If identifies any problematic resources that seem relevant to your situation, look for a details link.

If you see any status other than “System is operating normally”, look closely to see whether this resource affects you.

Click the details link, read it carefully, and follow any relevant instructions offered there.

If the details are consistent with your situation, follow instructions there before contacting Support.

Test 2: Try performing the build more than once before being sure of a problem.

Diagnosis 2: If your first build attempt fails but a later build attempt succeeds, this suggests that the build system was simply too busy during the failed attempt. This can occur intermittently for brief periods during peak loads; it does not predict future failures.

If multiple build attempts fail and no general status problems match your situation, contact Rackspace Support to ask for help.

Many Rackspace pages offer links to help you begin a live chat session or a telephone conversation with Support. Contact Rackspace also provides links to help you submit support tickets.

See also

To learn more about troubleshooting or preventing server build issues, read these articles:

  • RackConnect power users guide suggests using the Cloud Servers API to build multiple servers in bursts rather than singly.
  • Set up your first server shows how to use a software development kit (SDK) for the programming language of your choice to create a cloud server.
  • Actions for Cloud Servers identifies actions you can take after a server is built. It links to information about how to perform each action with the Cloud Control Panel and Cloud Servers API.

Getting support

You can find support for your work in the Rackspace cloud in many ways:

  • Support Home offers frequently-needed support and startup information.
  • System Status identifies alerts and disruptions.
  • Rackspace How-To publishes documentation primarily written for Cloud Control Panel users.
  • :rax-docs:`API documentation, including release notes, developer guides, and
    getting started guides <>`

Many Rackspace pages offer links to help you begin a live chat session or a telephone conversation. Contact Rackspace also provides links to help you submit support tickets.

Depending on which support tier you chose when creating your account, you might also have direct access to a specific account support team.

Trying something new

If you have never used a particular service before, your first attempt to do so in the Cloud Control Panel will remind you that you are beginning something new and point out the first step that you should take.

The Cloud Control Panel points out how to begin using a new service.

When Rackspace makes a new service generally available, all customers are given access to that service. When a new service is made available to you, its endpoint is added to the service catalog that is returned when you authenticate before using an API or CLI.

If you meet eligibility criteria, it might be possible for you to experiment with a new service before it is made available to all customers. The Rackspace blog is a good place to learn about limited-access and early-availability releases; signup procedures and contact information are included in each announcement.

Keeping up with change

Rackspace is growing and we frequently announce new features, new products, new prices, and other important changes. Here are some good ways to keep up with the news: