Win a Pass to Oracle OpenWorld

Want to win a pass to Oracle OpenWorld 2019 in San Francisco? Follow these steps to receive the giveaway:

1. Create a free Oracle Cloud trial account, if you don’t already have one.

2. Set up your tenancy.

3. Run a benchmark that we’ve documented or run your own benchmark to experience the performance for yourself.

4. Next, take a screenshot of your results and post it on Twitter.com using the hashtag #LetsProveItGiveaway

5. The first 20 eligible Oracle Cloud account holders to post the above Tweet will get one (1) pass to Oracle OpenWorld 2019 in San Francisco, a USD$2,050 value.


The first twenty (20) eligible Oracle Cloud account holders to perform a benchmark and post a screenshot of their results on Twitter.com using the hashtag #LetsProveItGiveaway will be given one (1) pass to Oracle Open World 2019 in San Francisco. CloudHarmony Cloud Network Test excluded. No purchase necessary. Must be 18 or older. Government employees, Oracle employees and their family members, and residents of Quebec, Italy, and countries under US embargo are not eligible to participate. Limit one per person. While supplies last.

Test Your Network Performance

Curious about network performance? In a matter of minutes, find out how Oracle Cloud Infrastructure compares to AWS and Azure downlink speeds and latency measurements.


Step 1: There’s nothing to download or launch. It’s as easy as clicking on the speed test link. We recommend benchmarking with Gartner CloudHarmony—not only is Cloud Harmony used in the following benchmarks, we use it ourselves to document performance.

Run a Block Storage Benchmark

Now that you have your Oracle tenancy, it’s time to run some benchmarks. Follow the step-by-step instructions below. Plan on spending about four hours on each benchmark from start to finish.


Block Storage Random Read/Write Benchmarks

Estimated time to complete: ~4 hours


Once your Oracle tenancy is ready, we recommend benchmarking with Gartner CloudHarmony. Not only is CloudHarmony used in the following benchmarks, we use it ourselves to document performance.

As one of the industry’s most comprehensive benchmarking suites for storage and other cloud services, CloudHarmony replicates the SNIA Solid State Storage (SSS) Performance Test Specification (PTS) Enterprise v1.1. You can find more details about the test specification here.

Before running any tests, protect and prevent any data loss. Do not run FIO tests directly against a device that is already in use, such as /dev/sdX. Run FIO only on unformatted raw devices that are not in use.

If the device is in use as a formatted disk with data on it, running FIO with a write workload (readwrite, randrw, write, trimwrite) will overwrite the data on the disk and cause data corruption. Proceed at your own risk.


Step 1: Find installation instructions for the Block Storage component of CloudHarmony here.


Step 2: Choose one of the block storage benchmarks below, or run your own. You can choose to benchmark configurations with standard or dense configurations.


In our testing, the benchmarks were run up to five consecutive times over several hours to simulate real-world situations and achieve steady-state results. Although time-consuming, we recommend the same practice.

The volume sizes range from 50 GB to 1 TB. 50 GB is the smallest size you can provision and a great fit for many applications with results of 3,000 IOPS. 1 TB delivers 25,000 IOPS and is ideal for applications with a small working set and large amounts of supporting data and applications on the same volume.


50 GB Block Volume

While on the smaller size, a 50 GB volume is great for starting out, especially when you’re sizing for capacity and don’t need a large block volume.

Test Details:
Block Volume Capacity: 50 GB
10GB working set
Direct I/O
Host Shape: Dense
Region: Phoenix

Reproduction Steps:
1. Mount 50 GB volume
2. Run Gartner Cloud Harmony Block Storage
a. ~/block-storage/run.sh --nopurge --noprecondition --fio_direct=1 --fio_size=10g --target /dev/sdb --test iops --skip_blocksize 512b

 

Observed Performance: 50 GB volume, 3,000 IOPS
Observed Performance: 50 GB volume, 3,000 IOPS graph

1 TB Block Volume

Test Details:
Block Volume Capacity: 1 TB
10GB working set
Direct I/O
Host Shape: Dense
Region: Phoenix

Reproduction Steps:
1. Mount 1 TB volume
2. Run Gartner Cloud Harmony Block Storage
a. ~/block-storage/run.sh --nopurge --noprecondition --fio_direct\=1 --fio_size=10g --target /dev/sdb --test iops --skip_blocksize 512b

 

Observed Performance: 1 TB, 25,000 IOPS
Observed Performance: 1 TB, 25,000 IOPS graph

20 x 1 TB Block Volumes

When 20 1 TB volumes are concurrently attached to the same dense instance with each volume tested independently, we see an incredible average of 2.1 GBPS and 400,000 IOPS to the host for the 50/50 read/write pattern.

Test Details:
Block Volume Capacity: 1 TB x 20
10GB working set
Direct I/O
Host Shape: Dense
Region: Ashburn

Reproduction Steps:
1. Mount x 20 1 TB volumes
2. Run Gartner Cloud Harmony Block Storage
a. ~/block-storage/run.sh --nopurge --noprecondition --fio_direct=1 --fio_size=10g --target /dev/sdb,/dev/sdc, /dev/sdd,/dev/sde,/dev/sdf,/dev/sdg,/dev/sdh,/dev/sdi, /dev/sdj,/dev/sdk,/dev/sdl,/dev/sdm,/dev/sdn,/dev/sdo, /dev/sdp,/dev/sdq,/dev/sdr,/dev/sds,/dev/sdt,/dev/sdu --test iops --skip_blocksize 512b

 

Observed Performance: 20 x 1 TB volumes, 400,000 IOPS
Observed Performance: 20 x 1 TB volumes, 400,000 IOPS graph

32 x 1 TB Block Volumes

When 32 volumes are concurrently attached to the same dense instance with each volume tested independently, we saw an incredible average of 400,000 IOPS to the host.

Test Details:
Block Volume Capacity: 32 x 1 TB
Direct I/O
Host Shape: Standard
Region: Ashburn

Reproduction Steps:
1. Provision 52 Core Standard2.52 or DenseIO2.52 BM
2. Mount x 32 1 TB volumes
3. Run Gartner Cloud Harmony Block Storage
a. ~/block-storage/run.sh   --target /dev/sdb,/dev/sdc, /dev/sdd,/dev/sde,/dev/sdf,/dev/sdg,/dev/sdh,/dev/sdi, /dev/sdj,/dev/sdk,/dev/sdl,/dev/sdm,/dev/sdn,/dev/sdo, /dev/sdp,/dev/sdq,/dev/sdr,/dev/sds,/dev/sdt,/dev/sdu, /dev/sdv,/dev/sdw,/dev/sdx,/dev/sdy,/dev/sdz,/dev/sdaa, /dev/ab,/dev/sdac,/dev/sdad,/dev/sdae,/dev/sdaf, /dev/sdag

 

Observed Performance: 32 x 1 TB volumes, > 400,000 IOPS
Observed Performance: 32 x 1 TB volumes, > 400,000 IOPS graph