VMware IOInsight is a tool to help people understand a VM's storage I/O behavior. By understanding their VM's I/O characteristics, customers can make better decisions at storage capacity planning and performance tuning. IOInsight ships as a virtual appliance that can be deployed in any vSphere environment and an intuitive web-based UI allows users to choose VMDKs to monitor and view results.
Where does IOInsight help?
- Customers may better tune and size their storage
- When contacting VMware Support for any vSphere storage issues, including a report from IOInsight can help VMware Support better understand the issues and can potentially lead to faster resolutions.
- VMware Engineering can optimize products with better understanding of various customers' application behavior.
IOInsight captures I/O traces from ESXi and generates various aggregated metrics that represent the I/O behavior. The IOInsight report contains only these aggregated metrics and there is no sensitive information about the application itself. In addition to the built-in metrics computed by IOInsight, users can also write new analyzer plugins to IOInsight and visualize the results. A comprehensive SDK and development guide is included in the download bundle.
- 4vCPU and 2GB VM memory
- VM storage space: 2GB (can grow up to 16GB)
- vSphere 5.5 or above
- Recommended browsers: Chrome and Safari
- SSH should be enabled in ESX machines to be monitored
Detailed instructions are provided in the VMware IOinsight User Guide.pdf
IOInsight v1.1.1 - April 12, 2017
- Added option to download logs for troubleshooting
- Removed 'Send Logs' option as we were not receiving mails from external networks due to mailserver issue
- Fixed support for ESXi-5.5 which was broken in v1.1.0
- Minor changes in UI and logs
- Commandline utility to setup static IP / DHCP
- Option to set NTP servers
- Option in UI to send logs and results to IOInsight developers for troubleshooting
- Fixed VC login issues
- Handling of several special characters in VM name
- Improved correctness of IO-size histogram
- Better handling of monitoring VMDKs with very less I/Os
- Improved alert messages in UI and logs
- Improved graph visualizations