Home > Storage > PowerScale (Isilon) > Industry Solutions and Verticals > Electronic Design Automation > EDA Cloud Burst with Dell PowerScale and Vcinity Data Access > Test method
Using the Synthetic EDA workload benchmark, 48 and 96 Android build jobs will be submitted using job scheduler through emulated WAN latency environments.
On-premises NFS mount has all files required to run Synthetic EDA workload benchmark. The Cloud NFS mount will only be used for working directories of each job to store source code, temporary files, and the final Android image compiled.
Vcinity plays a key role here to create a highly efficient RDMA connection between on-premises NFS shares and public cloud storages. As data is read through the Vcinity tunnel for the first time, it is stored in cache at the destination site. Cached files can be accessed instantly and repeatedly.
The local mirror copy of Google’s Android Open-Source Project is placed on a PowerScale cluster. The Android dataset contains around 1M file count, mostly small files, and a total size of 1.5TB. During the Synthetic EDA workload benchmark, each job will sync the main branch repo to its working directory on cloud NFS mount.
The EDA data is accessed over Vcinity’s highly efficient RDMA connection only. Emulated WAN latency values of 30, 60, and 120ms RTT are used to simulate distance and latency between on-premises and public cloud.
The EDA data is accessed over Vcinity’s highly efficient RDMA connection with the data cached on the public cloud (destination) side. Emulated WAN latency values of 30, 60, and 120ms RTT are used to simulate distance and latency between on-premises and public cloud.