developed our first cloud benchmark, CloudXPRT,
to measure the performance of cloud applications deployed on modern infrastructure
as a service (IaaS) platforms. When we first released CloudXPRT in
February of 2021, the benchmark included two test packages: a web microservices
workload and a data analytics workload. Both supported on-premises and cloud
service provider (CSP) testing with Amazon Web Services (AWS), Google Cloud
Platform (GCP), and Microsoft Azure.
is our most complex benchmark, requiring sustained compatibility between many
software components across multiple independent test environments. As vendors
roll out updates for some components and stop supporting others, it’s
inevitable that something will break. Since CloudXPRT’s launch, we’ve become
aware of installation failures while attempting to set up CloudXPRT on Ubuntu
virtual machines with GCP and Microsoft Azure. Additionally, while the web
microservices workload continues to run in most instances with a few
configuration tweaks and workarounds, the data analytics workload fails
consistently due to compatibility issues with Minio, Prometheus, and Kafka
within the Kubernetes environment.
response, we’re working to fix problems with the web microservices workload and
bring all necessary components up to date. We’re developing an updated test
package that will work on Ubuntu 22.04, using Kubernetes v1.23.7 and Kubespray
v2.18.1. We’re also updating Kubernetes Metrics Server from v1beta1 to v1, and will
incorporate some minor script changes. Our goal is to ensure successful
installation and testing with the on-premises and CSP platforms that we
supported when we first launched CloudXPRT.
are currently focusing on the web microservices workload for two reasons.
First, more users have downloaded it than the data analytics workload. Second, we
think we have a clear path to success. Our plan is to publish the updated web
microservices test package, and see what feedback and interest we receive from
users about a possible data analytics refresh. The existing data analytics workload
will remain available via CloudXPRT.com for the time being to serve as a
apologize for the inconvenience that these issues have caused. We’ll provide
more information about a release timeline and final test package details here
in the blog as we get closer to publication. If you have any questions about
the future of CloudXPRT, please feel free to contact us!
recently published a set of CloudXPRT Data Analytics and Web Microservices
workload test results
submitted by Quanta Computer, Inc.
The Quanta submission is the first set of CloudXPRT results that we’ve
published using the formal results submission and approval process.
We’re grateful to the Quanta team for carefully following the submission
guidelines, enabling us to complete the review process without a hitch.
If you are unfamiliar
with the process, you can find general information about how we review
submissions in a previous blog post.
Detailed, step-by-step instructions are available on the results submission page.
As a reminder for testers who are considering submitting results for July, the
submission deadline is tomorrow, Friday July 16, and the publication date is
Friday July 30. We list the submission and publication dates for the rest of
2021 below. Please note that we do not plan to review submissions in December,
so if we receive results submissions after November 30, we may not publish them
until the end of January 2022.
Submission deadline: Tuesday 8/17/21
Publication date: Tuesday 8/31/21
Submission deadline: Thursday 9/16/21
Publication date: Thursday 9/30/21
Submission deadline: Friday 10/15/21
Publication date: Friday 10/29/21
Submission deadline: Tuesday 11/16/21
Publication date: Tuesday 11/30/21
Submission deadline: N/A
Publication date: N/A
If you have any questions about the CloudXPRT results submission, review, or publication process, please let us know!
We recently received questions about whether we accept CloudXPRT
results submissions from testing on pre-production gear, and how we would handle
any differences between results from pre-production and production-level tests.
To answer first question, we are not opposed to pre-production
results submissions. We realize that vendors often want to include benchmark
results in launch-oriented marketing materials they release before their
hardware or software is publicly available. To help them do so, we’re happy to
consider pre-production submissions on a case-by-case basis. All such submissions
must follow the normal CloudXPRT results
submission process, and undergo
vetting by the CloudXPRT Results Review Group according to the standard review
and publication schedule. If we decide to publish pre-production results on our site, we
will clearly note their pre-production status.
In response to the second question, the CloudXPRT Results Review Group will handle any challenges to published results or perceived discrepancies between pre-production and production-level results on a case-by-case basis. We do not currently have a formal process for challenges; anyone who would like to initiate a challenge or express comments or concerns about a result should address the review group via firstname.lastname@example.org. Our primary concern is always to ensure that published results accurately reflect the performance characteristics of production-level hardware and software. If it becomes necessary to develop more policies in the future, we’ll do so, but we want to keep things as simple as possible.
If you have any questions about the CloudXPRT results submission process, please let us know!
Over the past few
weeks, we’ve received questions about whether we require specific test
configuration settings for official CloudXPRT results submissions. Currently, testers have the option to edit up to 12 configuration
options for the web microservices workload and three configuration options for the
data analytics workload. Not all configuration options have an impact on
testing and results, but a few of them can drastically affect key results
metrics and how long it takes to complete a test. Because new CloudXPRT testers
may not anticipate those outcomes, and so many configuration permutations are
possible, we’ve come up with a set of requirements for all future results
submissions to our site. Please note that testers are still free to adjust all
available configuration options—and define service level agreement (SLA)
settings—as they see fit for their own purposes. The requirements below apply only
to results testers want to submit for publication consideration on our site,
and to any resulting comparisons.
results submission requirement
Starting with the May results
submission cycle, all web microservices results submissions must have the workload.cpurequestsvalue, which lets the user designate the number of CPU cores the workload
assigns to each pod, set to 4. Currently, the benchmark supports values of 1,
2, and 4, with the default value of 4. While 1 and 2 CPU cores per pod may be
more appropriate for relatively low-end systems or configurations with few
vCPUs, a value of 4 is appropriate for most datacenter processors, and it often
enables CSP instances to operate within the benchmark’s max default 95th
percentile latency SLA of 3,000 milliseconds.
In future CloudXPRT releases, we may remove the option to change the workload.cpurequests value from the config.json file and simply fix the value in the benchmark’s code to promote test predictability and reasonable comparisons. For more information about configuration options for the web microservices workload, please consult the Overview of the CloudXPRT Web Microservices Workload white paper.
Data analytics results
Starting with the May
results submission cycle, all data analytics results submissions must have the best
reported performance (throughput_jobs/min) correspond to a 95th
percentile SLA latency of 90 seconds or less. We have received submissions where
the throughput was extremely high, but the 95th percentile SLA
latency was up to 10 times the 90 seconds that we recommend in CloudXPRT
documentation. High latency values may be acceptable for the unique purposes of
individual testers, but they do not provide a good basis for comparison between
clusters under test. For more information about configuration options with the
data analytics workload, please consult the Overview of the CloudXPRT Data Analytics Workload white paper.
We will update
CloudXPRT documentation to make sure that testers know to use the default
configuration settings if they plan to submit results for publication. If you
have any questions about CloudXPRT or the CloudXPRT results submission process,
please let us know.
We’re happy to announce
that CloudXPRT v1.1 will move from beta to general release status tomorrow! The
installation packages will be available at the CloudXPRT.com download page and the BenchmarkXPRT GitHub repository. You will find more details about the v1.1
updates in a previous blog post, but the most
prominent changes are the consolidation of the five previous installation
packages into two packages (one per workload) and added support for Ubuntu
20.04.2 with on-premises testing.
Before you get started
with v1.1, please note the following updated system requirements:
- Ubuntu 20.04.2 or later for on-premises testing
- Ubuntu 18.04 and 20.04.2 or later for CSP (AWS/Azure/GCP)
CloudXPRT is designed
to run on high-end servers. Physical nodes or VMs under test must meet the
following minimum specifications:
- 16 logical or virtual CPUs
- 8 GB of RAM
- 10 GB of available disk space (50 GB for the data analytics
We have also made
significant adjustments to the installation and test configuration instructions
in the readmes for both workloads, so please revisit these documents even if
you’re familiar with previous test processes.
As we noted during the
beta period, we have not observed any significant differences in performance
between v1.01 and v1.1, but we haven’t tested every possible test configuration
across every platform. If you observe different results when testing the same
configuration with v1.01 and v1.1, please send us the details so we can
If you have any questions about CloudXPRT v1.1, please let us know!
Last week, we announced that a CloudXPRT v1.1
beta was on the way. We’re happy to say that the v1.1 beta is now available to
the public on a dedicated CloudXPRT v1.1 beta download page. While CloudXPRT v1.01
remains the officially supported version on CloudXPRT.com and in our GitHub
repository, interested testers can use the v1.1
beta version in new environments as we finalize the v1.1 build for official
release. You are welcome to publish results as we do not expect results to
change in the final, official release.
As we mentioned in
last week’s post, the CloudXPRT v1.1 beta includes the following changes:
- We’ve added support for Ubuntu 20.04.2 or later for on-premises
- We’ve consolidated and standardized the installation packages
for both workloads. Instead of one package for the data analytics workload and
four separate packages for the web microservices workload, each workload has a
single installation package that supports on-premises testing and testing with
all three supported CSPs.
- We’ve incorporated Terraform to help create and
configure VMs, which helps to prevent problems when testers do not allocate
enough storage per VM prior to testing.
- We’ve replaced the Calico network plugin in Kubespray with Weave, which helps to avoid some
of the network issues testers have occasionally encountered in the CPS
Please feel free to
share the link to the beta download page. (To avoid confusion, the beta will
not appear in the main CloudXPRT download table.) We can’t yet state
definitively whether results from the new version will be comparable to those
from v1.01. We have not observed any significant differences in performance,
but we haven’t tested every possible test configuration across every platform.
If you observe different results when testing the same configuration with v1.01
and v1.1 beta, please send us the details so we can investigate.
If you have any questions about CloudXPRT or the CloudXPRT v1.1 beta, please let us know!