Disk Activity - VM Test
This test reports statistics pertaining to the input/output utilization of each physical disk on a guest.
Target of the test : A XenServer host
Agent deploying the test : An internal/remote agent
Outputs of the test : One set of results for every combination of virtual_guest:disk_partition
Configurable parameters for the test
|
Measurements made by the test
Measurement | Description | Measurement Unit | Interpretation |
---|---|---|---|
Percent virtual disk busy: |
Indicates the percentage of elapsed time during which the disk is busy processing requests (i.e., reads or writes). |
Percent |
Comparing the percentage of time that the different disks are busy, an administrator can determine whether load is properly balanced across the different disks. |
Percent reads from virtual disk: |
Indicates the percentage of elapsed time that the selected disk drive is busy servicing read requests. |
Percent |
|
Percent writes to virtual disk: |
Indicates the percentage of elapsed time that the selected disk drive is busy servicing write requests. |
Percent |
|
Virtual disk read time: |
Indicates the average time in seconds of a read of data from the disk. |
Secs |
|
Virtual disk write time: |
Indicates the average time in seconds of a write of data from the disk. |
Secs |
|
Avg. queue for virtual disk: |
Indicates the average number of both read and write requests that were queued for the selected disk during the sample interval. |
Number |
|
Current queue for virtual disk: |
The number of requests outstanding on the disk at the time the performance data is collected. |
Number |
This measure includes requests in service at the time of the snapshot. This is an instantaneous length, not an average over the time interval. Multi-spindle disk devices can have multiple requests active at one time, but other concurrent requests are awaiting service. This counter might reflect a transitory high or low queue length, but if there is a sustained load on the disk drive, it is likely that this will be consistently high. Requests experience delays proportional to the length of this queue minus the number of spindles on the disks. This difference should average less than two for good performance. |
Reads from virtual disk: |
Indicates the number of reads happening on a logical disk per second. |
Reads/Sec |
A dramatic increase in this value may be indicative of an I/O bottleneck on the guest. |
Data reads from virtual disk: |
Indicates the rate at which bytes are transferred from the disk during read operations. |
KB/Sec |
A very high value indicates an I/O bottleneck on the guest. |
Writes to virtual disk: |
Indicates the number of writes happening on a local disk per second. |
Writes/Sec |
A dramatic increase in this value may be indicative of an I/O bottleneck on the guest. |
Data writes to virtual disk: |
Indicates the rate at which bytes are transferred from the disk during write operations. |
KB/Sec |
A very high value indicates an I/O bottleneck on the guest. |
Disk service time: |
Indicates the average time that this disk took to service each transfer request ( i.e., the average I/O operation time) |
Secs |
A sudden rise in the value of this measure can be attributed to a large amount of information being input or output. A consistent increase however, could indicate an I/O processing bottleneck. |
Disk queue time: |
Indicates the average time that transfer requests waited idly on queue for this disk. |
Secs |
Ideally, the value of this measure should be low. |
Disk IO time: |
Indicates the average time taken for read and write operations of this disk. |
Secs |
The value of this measure is the sum of the values of the Disk service time and Disk queue time measures. A consistent increase in the value of this measure could indicate a latency in I/O processing. |
Avg IO read size: |
Indicates the average number of bytes transferred from disk during read operations. |
KB |
Larger I/Os tend to have higher latency (for example, BACKUP/RESTORE operations issue 1 MB transfers by default). These measures are reported for Windows VMs only. |
Avg I/O write size: |
Indicates the average number of bytes transferred into disk during write operations. |
KB |
|
Split IO: |
Reports the rate at which the operating system divides I/O requests to the disk into multiple requests. |
Splits/Sec |
A split I/O request might occur if the program requests data in a size that is too large to fit into a single request or if the disk is fragmented. Factors that influence the size of an I/O request can include application design, the file system, or drivers. A high rate of split I/O might not, in itself, represent a problem. However, on single-disk systems, a high rate for this counter tends to indicate disk fragmentation. This measure is reported for Windows VMs only. |
The detailed diagnosis of the Percent virtual disk busy measure, if enabled, provides information such as the Process IDs executing on the disk, the Process names, the rate at which I/O read and write requests were issued by each of the processes , and the rate at which data was read from and written into the disk by each of the processes. In the event of excessive disk activity, the details provided in the detailed diagnosis page will enable users to figure out which process is performing the I/O operation that is keeping the disk busy. The detailed diagnosis for this test is available for Windows guests only, and not Linux guests.
Figure 1 : The detailed diagnosis of the Percent virtual busy measure
Configuring Users for VM Monitoring
In order to enable the eG agent to connect to VMs in multiple domains and pull out metrics from them, the eG administrative interface provides a special page using which the different domain names, and their corresponding admin user names and admin passwords can be specified. To access this page, just click on the Click here hyperlink in any of the VM test configuration pages.
Figure 2 : Configuring a VM test
Upon clicking, Figure 3 will appear, using which the VM user details can be configured.
Figure 3 : The VM user configuration page
To add a user specification, do the following:
- First, provide the name of the Domain to which the VMs belong (see Figure 3). If one/more VMs do not belong to any domain, then, specify none here.
- The eG agent must be configured with user privileges that will allow the agent to communicate with the VMs in a particular domain and extract statistics. If none is specified against Domain, then a local user account can be provided against Admin User. On the other hand, if a valid Domain name has been specified, then a domain administrator account can be provided in the Admin User text box. If key-based authentication is implemented between the eG agent and the SSH daemon of a Linux guest, then, in the admin user text box, enter the name of the user whose <user_home_dir> (on that Linux guest) contains a .ssh directory with the public key file named authorized_keys. The admin password in this case will be the passphrase of the public key; the default public key file that is bundled with the eG agent takes the password eginnovations. Specify this as the admin password if you are using the default private/public key pair that is bundled with the eG agent to implement key-based authentication. On the other hand, if you are generating a new public/private key pair for this purpose, then use the passphrase that you provide while generating the pair. For the detailed procedure on Implementing Key-based Authentication refer to of this document.
- The password of the specified Admin User should be mentioned in the Admin Pwd text box.
- Confirm the password by retyping it in the Confirm Pwd text box.
-
To add more users, click on the
button in Figure 3. This will allow you to add one more user specification as depicted by Figure 4.
-
In some virtualized environments, the same Domain could be accessed using multiple Admin User names. For instance, to login to a Domain named egitlab, the eG agent can use the Admin User name labadmin or the Admin User name jadmn. You can configure the eG agent with the credentials of both these users as shown by Figure 5.
Figure 5 : Associating a single domain with different admin users
When this is done, then, while attempting to connect to the domain, the eG agent will begin by using the first Admin User name of the specification. In the case of Figure 5, this will be labadmin. If, for some reason, the agent is unable to login using the first Admin User name, then it will try to login again, but this time using the second Admin User name of the specification - i.e., jadmn in our example (see Figure 5). If the first login attempt itself is successful, then the agent will ignore the second Admin User name.
- To clear all the user specifications, simply click the Clear button in Figure 5.
- To remove the details of a particular user alone, just click the
button in Figure 5.
- To save the specification, just click on the Update button in Figure 5. This will lead you back to the test configuration page, where you will find the multiple domain names, user names, and passwords listed against the respective fields (see Figure 5).
Figure 6 : The test configuration page displaying multiple domain names, user names, and passwords