• Save
STG301 Using Amazon Elastic Block Store - AWS re: Invent 2012
Upcoming SlideShare
Loading in...5
×
 

STG301 Using Amazon Elastic Block Store - AWS re: Invent 2012

on

  • 1,677 views

This session will walk through best practices on architecting Amazon EBS for high performance workloads.

This session will walk through best practices on architecting Amazon EBS for high performance workloads.

Statistics

Views

Total Views
1,677
Views on SlideShare
1,677
Embed Views
0

Actions

Likes
2
Downloads
0
Comments
0

0 Embeds 0

No embeds

Accessibility

Categories

Upload Details

Uploaded via as Adobe PDF

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Processing…
Post Comment
Edit your comment

    STG301 Using Amazon Elastic Block Store - AWS re: Invent 2012 STG301 Using Amazon Elastic Block Store - AWS re: Invent 2012 Presentation Transcript

    • © 2012 Amazon.com, Inc. and its affiliates. All rights reserved. May not be copied, modified or distributed in whole or in part without the express consent of Amazon.com, Inc.
    • © 2012 Amazon.com, Inc. and its affiliates. All rights reserved. May not be copied, modified or distributed in whole or in part without the express consent of Amazon.com, Inc.
    • © 2012 Amazon.com, Inc. and its affiliates. All rights reserved. May not be copied, modified or distributed in whole or in part without the express consent of Amazon.com, Inc.
    • © 2012 Amazon.com, Inc. and its affiliates. All rights reserved. May not be copied, modified or distributed in whole or in part without the express consent of Amazon.com, Inc.
    • http://www.amazon.com/Seagate-Barracuda-3-5-Inch-Internal-ST3000DM001/dp/B005T3GRLY/ref=sr_1_2?s=pc&ie=UTF8&qid=1350499502&sr=1-2&keywords=hard+drive © 2012 Amazon.com, Inc. and its affiliates. All rights reserved. May not be copied, modified or distributed in whole or in part without the express consent of Amazon.com, Inc.
    • AmazonProvisioned IOPS Elastic Block Storage (EBS) IOPS: Within 10% of up to 2000 IOPS, 99.9% of the time, as provisioned. Throughput: 16k per IO = up to 32mbp/s, as provisioned. Latency: Each IO has a service time of provisioned IOPS/s Capacity: As provisioned, up to 1TB * IOPS: ~100 IOPS steady-state, with best-effort bursts to hundreds Throughput: variable by workload, best effort to 10’s of mbp/s. Latency: Varies, reads typically <20ms writes typically <10msAmazon Elastic Capacity: As provisioned, up to 1TB Block Storage (EBS) © 2012 Amazon.com, Inc. and its affiliates. All rights reserved. May not be copied, modified or distributed in whole or in part without the express consent of Amazon.com, Inc.
    • * Amazon Elastic Compute Cloud (EC2) Amazon Elastic Compute Cloud (EC2) Amazon Elastic Block Storage (EBS) Amazon Elastic Block Storage (EBS) © 2012 Amazon.com, Inc. and its affiliates. All rights reserved. May not be copied, modified or distributed in whole or in part without the express consent of Amazon.com, Inc.
    • TIME FOR SOME PROOF © 2012 Amazon.com, Inc. and its affiliates. All rights reserved. May not be copied, modified or distributed in whole or in part without the express consent of Amazon.com, Inc.
    • © 2012 Amazon.com, Inc. and its affiliates. All rights reserved. May not be copied, modified or distributed in whole or in part without the express consent of Amazon.com, Inc.
    • Benchmarks in the Cloud Steve Newman Scalyr
    • How Ill Spend Seven Minutes• Engineering decisions require data• Good decisions require good data• In the cloud, that means multiple measurements
    • Importance of Perf Benchmarks• Validate design assumptions• Evaluate alternatives• Capacity planning / load tests
    • Controlled Experiments Workload Application Application Server Type 1 Server Type 2 103.8 152.9
    • What Are You Measuring?• Behavior of an application• With a particular tuning• Under some workload• On a specific server type• And system configuration• From a particular hosting provider• In a particular data center• On a specific machine• At a point in time• Under some state of the network and supporting services
    • Some Variables are Outside Your Control
    • Test Many, Trust None• Run each test many times o On different instances o At different times• Aggregating across instances eliminates instance variability• Comparing instances measures instance variability
    • Wrapup• Further reading -- EC2 I/O analysis:http://blog.scalyr.com/2012/10/16/a- systematic-look-at-ec2-io/• Questions / comments:steve@scalyr.com
    • © 2012 Amazon.com, Inc. and its affiliates. All rights reserved. May not be copied, modified or distributed in whole or in part without the express consent of Amazon.com, Inc.
    • Miles Ward – Solutions Architect - @milesward – miward@amazon.com © 2012 Amazon.com, Inc. and its affiliates. All rights reserved. May not be copied, modified or distributed in whole or in part without the express consent of Amazon.com, Inc.
    • We are sincerely eager tohear your FEEDBACK on thispresentation and on re:Invent. Please fill out an evaluation form when you have a chance. © 2012 Amazon.com, Inc. and its affiliates. All rights reserved. May not be copied, modified or distributed in whole or in part without the express consent of Amazon.com, Inc.