First of all, what is VMware View Planner?
VMware View Planner is a comprehensive standard methodology for comparing virtual desktop deployment platforms. Using patented technology, View Planner generates a realistic measure of client-side desktop performance for all desktops being measured on the virtual desktop platform. View Planner uses a rich set of commonly used applications as the desktop workload.
View Planner provides a consistent methodology that captures the overall scalability of the deployment platform as well as the performance of each individual application operation within a given desktop. The number of virtual desktops for which workloads have met a defined performance threshold determines the overall benchmark score.
The benchmark utilizes VMware vSphere and VMware View as the virtualization infrastructure running on hardware infrastructure of choice, and is scalable from a few virtual machines running on one VMware vSphere host up to hundreds of virtual machines distributed across a cluster of vSphere hosts.
View Planner Benchmark Scores
When View Planner is used as a benchmark, the configuration of the workload virtual machines, the workload mix, and the versions of the View Planner controller appliance, operating systems, tools, and all other software used must conform to the specifications in the View Planner documentation (both this VMware View Planner Installation and User Guide and the View Planner Run and Reporting Rules).
To generate a VDImark™ benchmark score, the 95th percentile of the Group A QoS results and the 95th percentile of the Group B QoS results during the three iterations in the steady-state phase of a View Planner run must each be at or below the default thresholds.
The View Planner VDImark benchmark score is the number of concurrent users (that is, the number of View Planner desktop virtual machines or RDSH sessions) in a compliant run. The View Planner benchmark score may be used only as detailed in the View Planner Run and Reporting Rules.
View Planner VDImark Benchmark Run Requirements:
In order for a View Planner benchmark run to be compliant, it must meet the following requirements:
- It must comply with all requirements listed in the View Planner Run and Reporting Rules.
- It must use remote mode.
- It must use linked clones (rather than full clones).
- The benchmark workload profile (StandardBenchmarkProfile_5i) must be selected.
- All applications must be selected.
- The Multimedia Application speed must be set to Fast.
- It must use the PCoIP display protocol.
- The screen resolution must be set to 1920×1200.
- It must contain exactly five iterations (including Ramp-Up and Ramp-Down).
- The ramp-up time must be no greater than 60 seconds plus 2 seconds per virtual machine in the test and must be no greater than 600 seconds.
- The 95th percentile of the Group A response times during steady-state must be 1.0 seconds or less.
- The 95th percentile of the Group B response times during steady-state must be 6.0 seconds or less.
- The compliance log should contain no violations.
Additional details regarding what constitutes a compliant benchmark run can be found in the View Planner Run and Reporting Rules.
Most of the above list is not that much to worry about but the “The compliance log should contain no violations” can be a tricky one if you try to benchmark your own desktop template! This log is generated after each run and can be found in the run directory on the appliance.
2017-02-08 01:32:33,986 INFO — Checking for compliance —
2017-02-08 01:32:33,986 INFO Checking in the AD group: VDI
2017-02-08 01:32:33,986 INFO Checking for thinktime : OK
2017-02-08 01:32:33,986 INFO Checking for iterations : OK
2017-02-08 01:32:33,986 INFO Checking for applist (WORD,EXCEL_SORT,ACROBAT,IE_APACHEDOC,7ZIP,PPT,FIREFOX,OUTLOOK,IE_WEBALBUM,VIDEO) : Not OK (WORD,EXCEL_SORT,IE_APACHEDOC,7ZIP,PPT,OUTLOOK,IE_WEBALBUM,VIDEO,UEBENCH)
2017-02-08 01:32:33,986 INFO Custom App present in the run
2017-02-08 01:32:33,986 INFO Video (FAST) selected : YES
2017-02-08 01:32:33,987 INFO Checking for protocol (PCoIP) : OK
2017-02-08 01:32:33,987 INFO Checking for mode (remote) : OK
2017-02-08 01:32:33,987 INFO Checking for ramp-up time <= minimum(2*num_vms+60, 600): OK
2017-02-08 01:32:33,987 INFO Checking for concurrent logons (64) : OK
2017-02-08 01:32:33,987 INFO Checking compliance for vm: TestVM
2017-02-08 01:32:34,040 INFO Checking for no. of vcpu (1) : Not OK (2)
2017-02-08 01:32:34,041 INFO Checking for virtual memory allocated (1024 MB): Not OK (4096 MB)
2017-02-08 01:32:34,041 INFO Checking for guest os : microsoft windows 7 (64-bit) – Not OK
2017-02-08 01:32:34,041 INFO Office version (2010) : Not OK
2017-02-08 01:32:34,041 INFO Adobe Reader Version ([‘9′, ’10’]) : Not OK (NA)
2017-02-08 01:32:34,041 INFO Internet Explorer Version (9) : Not OK (8)
2017-02-08 01:32:34,041 INFO Screen Resolution (1920 x 1200) : Not OK (1626 x 817)
2017-02-08 01:32:34,041 WARNING The compliance failed for this VM: TestVM (skipping..)
2017-02-08 01:32:34,041 WARNING ==== There were 8 compliance violations. Please don’t worry about this compliance warnings unless you want to use “VDImark” term for publishing purposes. =====
As you see above, in this run there are a few compliancy issues that block the VDImark score being present in the report and is set to 0.
1. Wrong benchmark profile is used, needs to be StandardBenchmarkProfile_5i.
2. VM is configured with 2 vCPU instead of 1.
3. VM is configured with 4GB ram instead of 1GB.
4. VM Guest OS is not correct, only supported is Windows 7 32 Bit with SP1.
5. Office Version is not correct, only supported is Microsoft Office Professional Plus 2010, 32-bit (with no service packs).
6. Adobe reader is (NA) due to installed version of Reader DC. Only supported is v9 to get VDImark.
7. Internet explorer version installed in VM is 8, only supported version is 9.
8. Screen resolution of client VM is not correct, must be 1920×1200.
Once you have created the correct template for the VDImark scoring via the user manual, then you should be able to get that VDImark score in the report. VDImark is very strict in it’s compliancies.
Be aware that you need to separate your client VM and desktop VM on separate clusters. Otherwise the client VM will impact the desktop VM during benchmarks. Also if you do not split those systems you will have less resources available for your desktop VM ergo lesser desktop in a run. In remote mode every desktop will have a unique client VM that initiates the connection.
Another side note: Most of you guys user OSOT (VMware OS Optimization Tool) this may break compliany! Happy benching!