In this session, get three takeaways about Perforce performance benchmarks and their results across varying storage protocols, using NetApp storage as an example. Learn how to use Perforce benchmarks and tools to validate the performance of your Perforce deployment; understand Perforce performance across different storage protocols; and get tips and tricks for deploying Perforce on varying storage technologies.
3 Ways to Improve Performance from a Storage Perspective
1. #
3 Ways to Improve Performance
from a Storage Perspective
Vibhor Gupta - NetApp
Tim Brazil - Perforce
2. #
Vibhor Gupta
Applications Eng.
NetApp Inc.
Tim Brazil
Performance Lab Eng.
Perforce Software
Tim Brazil entered the software industry as a Sys
Admin and soon moved into Quality Assurance
roles at leading database companies. He joined
Perforce in March 2007, where his focus is
performance-related issues as they apply to
client/server environments.
Vibhor Gupta is a Technical Marketing
Engineer at in Tech-Apps at NetApp. He
partnered with Perforce, working on the
characterization of the Perforce Application on
NetApp Storage. He also is involved in the
characterization of the EDA workload, where
Perforce is heavily used for SCM
3. #
Agenda
• Introduction
• Benchmarks
• Test Results and Analysis
• Key Takeaways
4. #
Introduction
• Perforce Server Deployment: Considerations
− Performance
− Cost
− Backup and Recovery
19. #
Best Practices
• For best write performance, Perforce metadata and journals should be
placed on faster storage like FC or iSCSI.
• For NFS, recommended mount options local_lock=all and nocto
• For heavy server logging, use FCP or iSCSI.
20. #
Key Takeaways
FC iSCSI iSCSI and NFS
(Split
Configuration)
NFS
Performance Excellent Excellent Good Fair
Cost Fair Excellent Excellent Excellent
Ease of backup and
recovery
Fair Fair Good Excellent
Protocol
Criterion
21. #
RESOURCES
Technical Report: http://www.netapp.com/us/media/tr-4164.pdf
Public Benchmarks: ftp://ftp.perforce.com/perforce/tools/benchmarks/
This illustrates the results from the browse benchmark. As you can see, FCP, iSCSI and split configuration the results were basically the same, however, for NFS, results were poor.
For the Deltas test, which is a read benchmark, FCP, iSCSI, and NFS are comparable in performance.
Summary of Best Practices include:
[first animation] For best write performance, Perforce metadata and journals should be placed on FC or iSCSI.
[second animation] If using NFS, recommended mount options include local_lock=all and nocto. (Mention it works only with dedicated NFS server).
[third animation] If the log level of Perforce application is set to a high value (3 or greater or debug) or the server activity introduces high logging, place logs on FC or iSCSI to alleviate performance issue.