In this article we’ll describe the Windows Azure monitor written in C#. Windows Azure has a set of different services, but we’ll limit our article to only those services that we use for capturing performance counters. If you’re new to Windows Azure and want more information you can visit the Microsoft Azure website.
Windows Azure Compute
There are three types of scalable compute instances which you can run in the cloud.
· Web Role – provide a dedicated Internet Information Services (IIS) web-server used for hosting front-end web applications.
· Worker Role – can run asynchronous, long-running or perpetual tasks independent of user interaction or input.
· VM Role – Virtual Machine (VM) roles, now in Beta, enable you to deploy a custom Windows Server 2008 R2 (Enterprise or Standard) image to Windows Azure.
The current version of our Azure Monitor supports capturing performance counters for the Web Role or the Worker Role. In addition to the standard performance counters, Windows Azure allows for custom performance counters.
In our custom monitor, the performance counters are configured by calling the SetupCounters method in project HostWorkerRole. The sample code is of the Worker Role which is deployed to the Azure Compute Service.
Windows Azure Storage.
Windows Azure Storage provides secure and scalable storage services that are highly available and durable. The Storage service supports virtually all types of storage needs, from structured to unstructured data, NoSQL databases, and queues. You can find more information about this service here. Some of the most practical uses of Azure Storage are:
- BLOB (Binary Large Object) storage. BLOB Storage is the simplest way to store large amounts of unstructured text or binary data such as video, audio and images.
- Table storage is used by applications requiring storing large amounts of data storage that need additional structure. While a table stores structured data, it does not provide any way to represent relationships between the data, sometimes called a NoSQL database. You can use SQL Azure for relational database service on Windows Azure.
- Queue used for reliable, persistent messaging between applications. You can use Queues to transfer messages between applications or services in Windows Azure.
- Windows Azure Drive allows applications to mount a Blob formatted as a single volume NTFS VHD. You can move your VHDs between private and public clouds using Windows Azure Drive.
All of these storage types related to a Storage Account, which is created in the Windows Azure dashboard.
Windows Azure Storage Analytics performs logging and provides metrics data for a storage account. You can use this data to trace requests, analyze usage trends, and diagnose issues with your storage account. A detailed overview is available on MSDN.
Storage Analytics Metrics can report transaction statistics and capacity data for a storage account. All metrics data is stored in two tables per service: one table for transaction information, and another table for capacity information. Transaction information consists of request and response data. Capacity information consists of storage usage data. As of version 1.0 of Storage Analytics, capacity data will only be reported for the Blob service. The current implementation of our monitor application supports analytics for the Table service. In the future it can be expanded to support Blobs and Queues.
Azure Pricing Information
Windows Azure Diagnostic collects diagnostic data from instances and copies it to a Window Azure Storage account (either on blob and table storage). Those diagnostic data (such as log) can indeed help developer for the purpose of monitoring performance and tracing source of failure if exception occurs.
We’ll need to define what kind of log (IIS Logs, Crash Dumps, FREB Logs, Arbitrary log files, Performance Counters, Event Logs, etc.) to be collected and send to Windows Azure Storage either on-schedule-basis or on-demand.
However, if you are not careful defining your needs for diagnostic information, you could end up paying an unexpected high bill. At the time of this writing, the cost is as follows:
- $0.14 per GB stored per month based on the daily average
- $0.01 per 10,000 storage transactions
Some Figures for Illustration
Assuming the following figures:
- You have a few applications that require high processing power of 100 instances
- You apply five performance counter logs (Processor% Processor Time, MemoryAvailable Bytes, PhysicalDisk% Disk Time, Network Interface Connection: Bytes Total/sec, Processor Interrupts/sec)
- Perform a scheduled transfer with an interval of 5 seconds
- The instance will run 24 hours per day, 30 days per month
Given the above scenario, the total number of transactions comes to 259,200,000 each month (5 counters X 12 times X 60 min X 24 hours X 30 days X 100 instances), or $259/month.
Now what if you don’t really need that many counters every 5 seconds and reduce them to 3 counters and monitor it every 20 seconds? In this case it would be 3 counters X 3 times X 60 min X 24 hours X 30 days X 100 instances = 3,8880,000 transactions, or $38/month. Windows Azure Diagnostic is needed but using it improperly may be more expensive than you bargained for.
Storage Analytics is enabled by a storage account owner; it is not enabled by default. All metrics data is written by the services of a storage account. As a result, each write operation performed by Storage Analytics is billable. Additionally, the amount of storage used by metrics data is also billable.
The following actions performed by Storage Analytics are billable:
- Requests to create blobs for logging
- Requests to create table entities for metrics
If you have configured a data retention policy, you are not charged for delete transactions when Storage Analytics deletes old logging and metrics data. However, delete transactions from a client are billable.
The Mediator application makes it possible to retrieve performance counter values and Storage Account Table metrics Windows Azure. The following performance counters are included:
- \Processor(_Total)\% Processor Time
- \Memory\Available Bytes
- TotalRequests – the number of requests made to a storage service or the specified API operation. This number includes successful and failed requests, as well as requests which produced errors.
- TotalBillableRequests – the number of billable requests.
- Availability – the percentage of availability for the storage service or the specified API operation. Availability is calculated by taking the TotalBillableRequests value and dividing it by the number of applicable requests,including those that produced unexpected errors. All unexpected errors result in reduced availability for the storage service or the specified API operation.
Each metric or performance counter gets its own monitor on the Dashboard at www.monitis.com.
Windows Azure Storage Metrics allows you to track your aggregated storage usage for Blobs, Tables and Queues. The details include capacity, per service request summary, and per API level aggregates. The metrics information is useful to see an aggregate view of how a given storage account’s blobs, tables or queues are doing over time. It makes it very easy to see the types of errors that are occurring to help tune your system and diagnose problems and the ability to see daily trends of its usage. For example, the metrics data can be used to understand the request breakdown (by hour).
The application performs the role of Mediator between the Windows Azure Table Service and public API of monitis.com. To get started, the first step is to specify the API key to access REST service on www.monitis.com.
After logging in the Mediator will check if all required monitors are already created on the Monitis Dashboard. If they are not, you can use the create button to create all the required monitors.
Once the monitors have been created, the “Next” button will enabled and you will see all the monitors created be Mediator on the dashboard.
Next, Mediator needs your account information for Windows Azure Storage Account so we can access the performance and metrics data for monitor. If you don’t have it the account information, you can check “Use Default” and all values will fill with test credentials.
Click on “Apply” to check the credentials and test the connection to the Windows Azure Table Service. The applications will also check the existence of the table PerformanceCounter data. This table will automatically be created when the installation configures capturing of performance counters. For example, the table name for performance counters is “WADPerformanceCountersTable”.
After this is done, you will see a screen where you can configure the interval period in seconds used synchronize the data to the Monitis monitor. If there is actual data in the performance counters table, you should see a chart like this:
The optimal value for the current performance counters configuration is 180 seconds.
For the Storage Account Analytics you will see a section that allows you to specify the time period you want to synchronize between Azure and Monitis.
Select the time period and click “Sync”. After this is completed you’ll see the result of Storage Analytics and performance counters:
Performance counters and analytics metrics can work together. This means that you can start the Mediation process for performance counters and at the same time use the “Sync” button for the Storage Account Analytics.
Here are some links with more information: