Comments on: VLDB Keynote: Data Access Patterns in The Amazon.com Technology Platform http://innocuous.org/articles/2007/09/25/vldb-keynote-data-access-patterns-in-the-amazoncom-technology-platform/ Richard Tibbetts on Various Topics Thu, 20 Jan 2011 11:35:50 +0000 http://wordpress.org/?v=2.9.2 hourly 1 By: Richard Tibbetts http://innocuous.org/articles/2007/09/25/vldb-keynote-data-access-patterns-in-the-amazoncom-technology-platform/comment-page-1/#comment-2 Richard Tibbetts Wed, 26 Sep 2007 13:13:53 +0000 http://new.innocuous.org/index.php/2007/09/25/vldb-keynote-data-access-patterns-in-the-amazoncom-technology-platform/#comment-2 The presentation is not yet available online. I'm not sure if/when it will become so. I will try to update with a link if I find it. The presentation is not yet available online. I’m not sure if/when it will become so. I will try to update with a link if I find it.

]]> By: Chris Lesniewski-Laas http://innocuous.org/articles/2007/09/25/vldb-keynote-data-access-patterns-in-the-amazoncom-technology-platform/comment-page-1/#comment-3 Chris Lesniewski-Laas Wed, 26 Sep 2007 01:27:47 +0000 http://new.innocuous.org/index.php/2007/09/25/vldb-keynote-data-access-patterns-in-the-amazoncom-technology-platform/#comment-3 Is the keynote online anywhere? I agree about the core message, but I'm curious about some of the details. I'd guess that most failures at Amazon are basically fail-stop, but that the hard cases are the fail-stutter case or outputting garbage. Do you think we need Byzantine tolerance to deal with garbage-spewing nodes, or can the garbage be constrained to some non-malicious model? What kinds of non-strongly-consistent, non-eventually-consistent models do they use? What are some examples of queries/updates at Amazon that don't need strong consistency? I can think of some candidates, but my own experience using Amazon seems to present a consistent interface. For example, after I add an item to my Wish List, my Wish List is always immediately updated --- I never see a temporary blip due to "eventual consistency". When I make a purchase, it's a pretty good bet that they need to make a real-time (i.e. strongly consistent) validation with the credit-card processor, even if some of their internal systems are not strongly consistent due to failures. So that might be better characterized as some strongly consistent "hard state" combined with some squishily consistent "soft state". Is the keynote online anywhere? I agree about the core message, but I’m curious about some of the details.

I’d guess that most failures at Amazon are basically fail-stop, but that the hard cases are the fail-stutter case or outputting garbage. Do you think we need Byzantine tolerance to deal with garbage-spewing nodes, or can the garbage be constrained to some non-malicious model?

What kinds of non-strongly-consistent, non-eventually-consistent models do they use?

What are some examples of queries/updates at Amazon that don’t need strong consistency? I can think of some candidates, but my own experience using Amazon seems to present a consistent interface. For example, after I add an item to my Wish List, my Wish List is always immediately updated — I never see a temporary blip due to “eventual consistency”. When I make a purchase, it’s a pretty good bet that they need to make a real-time (i.e. strongly consistent) validation with the credit-card processor, even if some of their internal systems are not strongly consistent due to failures. So that might be better characterized as some strongly consistent “hard state” combined with some squishily consistent “soft state”.

]]>