TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Amazon Aurora Backtrack

416 pointsby jeffbarrabout 7 years ago

23 comments

ulkeshabout 7 years ago
&quot;Aurora will try to retain enough log information to support that window of time.&quot;<p>It&#x27;s good to know that Aurora will try. It&#x27;s not like it needs to be reliable or anything.
评论 #17039317 未加载
评论 #17039526 未加载
评论 #17039863 未加载
评论 #17042938 未加载
评论 #17044533 未加载
dugganabout 7 years ago
Aurora keeps coming along in leaps and bounds, congratulations to the team, this is a fantastic achievement!<p>I only wish that every new feature didn&#x27;t inevitably come with the caveat that it&#x27;s only for the MySQL flavour of Aurora.<p>I understand both the engineering and product development reasons for doing so (different stack and MySQL is undoubtedly a much larger customer base), but it always makes these announcements a little underwhelming as an Aurora Postgres user.
评论 #17039959 未加载
评论 #17039641 未加载
评论 #17044235 未加载
Bucephalus355about 7 years ago
Ok Oracle has had this feature for at least a decade, it’s called a “flashback query”. Obviously Aurora costs 10% of Oracle, but still, I thought this was going to be a huge feature-add considering the HN comment count.<p>That being said, I love AWS, am Pro-Certified, and work with it everyday.<p>I know Oracle is a giant mean bully company, but at least their arrogance was never of the “world-destabilizing” kind like Facebook.<p>EDIT: changed rollback query to flashback query (flashback query can be used both to view or to actually change the DB)
评论 #17047032 未加载
评论 #17040418 未加载
ben509about 7 years ago
Reading the paper[1] linked from Jeff&#x27;s post:<p>&gt; In Aurora, we have chosen a design point of tolerating (a) losing an entire AZ and one additional node (AZ+1) without losing data, and (b) losing an entire AZ without impacting the ability to write data. We achieve this by replicating each data item 6 ways across 3 AZs with 2 copies of each item in each AZ. We use a quorum model with 6 votes (V = 6), a write quorum of 4&#x2F;6 (V w = 4), and a read quorum of 3&#x2F;6 (V r = 3). With such a model, we can (a) lose a single AZ and one additional node (a failure of 3 nodes) without losing read availability, and (b) lose any two nodes, including a single AZ failure and maintain write availability. Ensuring read quorum enables us to rebuild write quorum by adding additional replica copies.<p>There are many 2 AZ regions in AWS, of course. I don&#x27;t think you can stripe 3 copies per AZ, an AZ failure drops you to potentially 2&#x2F;6, and if you allow for 2&#x2F;6 and 3&#x2F;6 writing you could have a split brain. Any thoughts how they manage that?<p>[1] <a href="https:&#x2F;&#x2F;www.allthingsdistributed.com&#x2F;files&#x2F;p1041-verbitski.pdf" rel="nofollow">https:&#x2F;&#x2F;www.allthingsdistributed.com&#x2F;files&#x2F;p1041-verbitski.p...</a>
评论 #17043599 未加载
评论 #17044385 未加载
评论 #17043645 未加载
gtsteveabout 7 years ago
This is nice but it appears that the entire database instance gets rolled back to that point. It&#x27;d be a lot nicer if it could be done at a per-db or per-table granularity.<p>Realistically I&#x27;d never use this feature because of the risk of data loss. I&#x27;d restore a new instance from backups and copy the lost data back over manually.
评论 #17042989 未加载
评论 #17042744 未加载
wpietriabout 7 years ago
Very interesting. The describe it as a rewind. Does anybody know if it&#x27;s really a rewind, where each log record is reversible? Or do they do the easier thing of saving snapshots and then replaying the log from snapshot to desired point?
评论 #17039163 未加载
评论 #17039111 未加载
评论 #17039044 未加载
thelastidiotabout 7 years ago
Amazon is the new IBM. Knock yourself out and jump into the AWS ecosystem. In a few years down the line, you&#x27;ll understand that you&#x27;ve lost the leverage you had to potentially take your public cloud business somewhere else when you have so many dependencies on Amazon tech. Basic principles from my view: don&#x27;t adopt anything but standard EC2&#x2F;S3 services and create diversity not only in your teams but in your infrastructure policies.
评论 #17040222 未加载
评论 #17040566 未加载
评论 #17042136 未加载
评论 #17040517 未加载
评论 #17042050 未加载
评论 #17040551 未加载
评论 #17039973 未加载
评论 #17042615 未加载
评论 #17040565 未加载
评论 #17041742 未加载
评论 #17043301 未加载
评论 #17056210 未加载
评论 #17044265 未加载
评论 #17040175 未加载
评论 #17041017 未加载
评论 #17039902 未加载
评论 #17040335 未加载
qiuyesuifengabout 7 years ago
TiDB has already supported this (similar) feature about 2 years ago and it has been adopted by the gaming users: <a href="https:&#x2F;&#x2F;www.pingcap.com&#x2F;blog&#x2F;2016-11-15-Travelling-Back-in-Time-and-Reclaiming-the-Lost-Treasures&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.pingcap.com&#x2F;blog&#x2F;2016-11-15-Travelling-Back-in-T...</a>
estsauverabout 7 years ago
I&#x27;m slightly confused, is this the same as the existing point-in-time restore that&#x27;s available for other RDS instances?<p>Edit: Main difference seems to be new cluster vs. in place.
评论 #17039277 未加载
评论 #17039114 未加载
评论 #17040362 未加载
评论 #17039115 未加载
评论 #17039297 未加载
aionicabout 7 years ago
It&#x27;s a relatively classic invention, take something that exists and repackage. A snapshot and a log replay accomplishes something pretty similar. AWS slapped a ui and some orchestration around it. The cloud lock stuff makes sense (although if having an easy &quot;undo button&quot; on your db layer is mission critical to your business you might have other interesteting challenges.
评论 #17079522 未加载
cody8295about 7 years ago
I don&#x27;t know anything about Aurora and maybe I&#x27;m missing something. But why not just wrap everything in a TRANSACTION and then do a ROLLBACK if there&#x27;s an issue?
评论 #17043399 未加载
评论 #17042266 未加载
评论 #17045306 未加载
评论 #17042179 未加载
craigkerstiensabout 7 years ago
At this time looks like it only applies to MySQL, will be curious to hear if&#x2F;when it becomes available for PostgreSQL.
setheronabout 7 years ago
How is this different than point in time restore already available?
评论 #17040888 未加载
评论 #17041007 未加载
评论 #17040656 未加载
manish_gillabout 7 years ago
The seamlessness of this feature is quite amazing. Backups are usually a huge pain to deal with (I&#x27;ve recently been dealing with Postgres&#x2F;Barman quite a bit). And disaster scenarios aside (for which AWS already does replications across regions), I think a frequent purpose of backups is really to do this &quot;Undo&quot;, go back in time and pretend something didn&#x27;t happen.<p>All this makes me really really wanna use Aurora. :)
评论 #17044684 未加载
rustywormabout 7 years ago
Interesting - but what if your database has constant activity? &quot;Oops, my SQL bad&quot; becomes &quot;Oops, my rewind lost 410 transactions&quot;?
评论 #17052726 未加载
评论 #17041947 未加载
polskibusabout 7 years ago
How did Aurora begin its life? Was it written from scratch or forked from existing open source database?
评论 #17042407 未加载
brettgo1982about 7 years ago
How is this better than their already existing PITR?<p>Why would someone want to rollback their own production database instead of PITR to a new database and switching over to it? Surely you would end up losing data because you wouldn&#x27;t be able to reconcile the new data written to it.
truth_seekerabout 7 years ago
CockroachDB also support this. <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=11958660" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=11958660</a>
评论 #17079542 未加载
sleepychuabout 7 years ago
&gt; <i>We’ve all been there! You need to make a quick, seemingly simple fix to an important production database.</i><p>Have we though? This could be one of those safety nets that makes me worse not better.
ethanpilabout 7 years ago
Does Amazon release the source for these features? Would love to see these ported to other flavors of mySQL.
评论 #17044818 未加载
truth_seekerabout 7 years ago
If you use Datomic with DynamoDB, this feature is available at Query level.
edge17about 7 years ago
I’m confused, how is this different than an Undo log?
评论 #17039491 未加载
qurasheeabout 7 years ago
Amazon rediscovering PITR, nice :P
评论 #17039550 未加载