We are excited to announce more self-service troubleshooting and diagnostics options to Support Portal for Azure Websites. Since it’s preview launch, we have added following new features.
1) Observe: Ability to monitor live HTTP traffic and server errors per hostname or combined for a given Azure website.
This scenario is very useful when your websites are behind traffic manager or accessed using different host names. This way you can monitor live traffic for a given region or specific hostname of your choice or all combined.
2) Analyze: Event Viewer – Ability to view event logs from all of your instances combined into single view.
You can chose to filter based on event levels, date-time range, event ID and Source. Ability to sort per instance is coming soon.
3) Analyze: Role Metrics – View various metrics per role instance for a given website (Only available for Basic and Standard Tier)
Your website can be running on multiple instances and at times it is very difficult to identify which instance is actually having a performance issue. Want to know – How much CPU, memory, average HTTP response time or requests served or network metrics is used by a given instance? – this is your place.
The following Metrics (For Last 5 days) have been exposed right now. Note that you can filter the graphs by clicking on corresponding metric circle.
a. CPU/Memory Metrics/Role
b. HTTP Metrics/Role
![clip_image001[4]](https://acom.azurecomcdn.net/80C57D/blogmedia/blogmedia/2014/11/26/clip_image0014.jpg "clip_image001[4]")
c. Network Metrics/Role
![clip_image001[6]](https://acom.azurecomcdn.net/80C57D/blogmedia/blogmedia/2014/11/26/clip_image0016.jpg "clip_image001[6]")
Now that you’ve identified an instance that needs an attention, you can analyze that instance using Diagnostics or simply Restart Site Process on a specific role instance instead of restarting the whole site and taking down all other instances at the same time.
Note: Upon clicking on restart W3WP, you are given 5 seconds time to cancel the operation, just in case you fat fingered it!
4) Analyze: ClearDB – Basic diagnostics data and tests for MySQL database provided by ClearDB
We partnered with ClearDB to have basic diagnostics information about any MySQL DB used by your subscription that you have logged on to support portal with. This data includes, information like: DB Tier, Size, Usage, Last Updates, Maximum Connections supported, Violations, Slow Logs, Open connections, Connection tests with it’s response time, ClearDB cluster state and over all DB health status.
5) Notifications: Azure Websites Support Portal will now alert you if we detect any http runtime issues with your websites.
We will try to detect 3 classes of issues, which are,
a) Application Issue: Your website is giving HTTP 5xx class of errors in past few minutes, while other vitals of the system/service are normal
b) Azure Websites Issue: Our internal monitoring has detected some service level degradation that might be impacting your website (this is something where we stop doing everything and work on resolving this issue)
c) Potential Azure Websites Issue: Our internal monitoring is detected an issue specific to machine or synthetic transaction that is specific to your site. This could be side effect of good application going bad or actual service issue isolated to just your site
As shown above, If you hover over an alert, the tool tip will show a detailed message on the type of the issue we detected. This alert notification rolls up to the subscription and “Select Website” level so it is easy to navigate to the site that is impacted at a given time.
Quick Refresher on how to access the support portal
How do I access this Site Extension? Simply browse your website’s SCM endpoint and select the Support option from the Tools menu.
Note: In case of an issue with Azure Websites service, above site extension might not be accessible via SCM endpoint. In that case, you can directly browse to following URLs and try and access the portal using azure credentials.
https://support-bay.scm.azurewebsites.net/support