Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
amix
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
55 ms
·
481.
▲
by
amix
18y ago
I would not recommend this if you want your product to hit the average user - IE has the biggest market share (still :(). But of course, for Instpaper it's perfectly right to do so since the market is power-users and they mostly us FF (as t
482.
▲
by
amix
18y ago
You get a taste of ice-cream - it's the equivalent of hearing a snippet of a song or seeing a trailer for a movie. You don't get a full banana split.
483.
▲
by
amix
18y ago
Their traffic is the triple of plentyoffish.com and their site is filled with advertising 1). I doubt it's that expensive to run a tracker and especially when they run it in Egypt. I can't really see how they could be broke - they probably
484.
▲
by
amix
18y ago
I think the Pirate Bay is a shady operation. They probably earn millions on giving other peoples work for free. I don't understand the model of "free copying" - if everyone is leeching and nobody is paying how will we get movies like Star W
485.
▲
by
amix
18y ago
Like I have already stated I am interested in how the system will run in production. Generally, you will do lots of small updates and lots of small fetches with key-value databases. You won't do batch operations - which makes your benchmark
486.
▲
by
amix
18y ago
memcachedb is not distributed - it only supports replication. I.e. with memcachedb you can only scale reads, but not writes (or at least not without a system like LightCloud on top of it).
487.
▲
by
amix
18y ago
These two servers run 3 lookup nodes and 6 storage nodes (i.e. 6 lookup nodes and 12 storage nodes in total). These servers are quite powerful [32GB of RAM and using RAID10], they also run MySQL.
488.
▲
by
amix
18y ago
Every node in both hash rings is replicated using master-master replication - i.e. node A and node A' can both receive updates and reads. Node A and node A' sync their updates via an update log and can fail at any time and come back at any
489.
▲
by
amix
18y ago
LightCloud adds horizontal scaling. If you just use Tokyo Tyrant then you can only scale by buying bigger servers. If you use LightCloud you can scale by buying extra servers. When scaling upwards you would generally _really_ want to scale
490.
▲
by
amix
18y ago
Please do see http://news.ycombinator.com/item?id=498699 (memcachedb != LightCloud). memcachedb should be compared with Tokyo Tyrant and not LightCloud. And if you liked, you could extend LightCloud with memcachedb support (which we also
491.
▲
by
amix
18y ago
At Plurk.com we use both LightCloud and memcached. They complement each other. I.e. memcached is used for caching stuff - to reduce the load to MySQL. LightCloud is used to store persistent data - such as how many times a plurk has been rea
492.
▲
by
amix
18y ago
memcachedb is not distributed, meaning that you can only scale vertically (i.e. by buying bigger machines). With LightCloud you can scale by adding extra servers to the system (i.e. horizontal scaling). This said, Tokyo Tyrant performs bett
493.
▲
by
amix
18y ago
memcached is a memory based key-value database. LightCloud is persistent i.e. data is saved to disk. I'll make this more clear on the website.
494.
▲
by
amix
18y ago
Comments, questions and reviews are very welcomed.
495.
▲
LightCloud: Distributed key-value database built on Tokyo Tyrant
(opensource.plurk.com)
62 points
by
amix
18y ago
|
26 comments
496.
▲
by
amix
18y ago
From my experience (and I have used Sphinx, Lucene, Solr and Xapian in production) is that Lucene/Solr have a pretty bad perfomance compared to Sphinx or Xapian. My Lucene setup began to throw deadlocks and memory exceptions pretty early on
497.
▲
by
amix
18y ago
I feel the same way - even thought I have created a task manager to solve these kind of issues (Todoist). The general problem for me is that I focus too much on one thing that I forget/postpone a lot of the other tasks and suddenly I have a
498.
▲
by
amix
18y ago
Why is this up rated? Why not just use tables and get the job done instead of "hacking" tables in CSS sigh
499.
▲
by
amix
18y ago
A simple tip for those that run any kind of database: Be sure to replicate them in master-slave (or master-master). And base your backups on taking a slave down for backups. Hot backups only work for very small databases - even those that a
500.
▲
by
amix
18y ago
For Tokyo Tyrant there's a Python library: http://code.google.com/p/pytyrant/ It uses the binary protocol and it's written in a very simple and very clean way - and it should be production ready. I would not recommend using the memcache
501.
▲
by
amix
18y ago
There's global locking and key locking. I currently use key-locking when updating. You should be able to implement CAS in Lua I guess, but it isn't supported from the start.
502.
▲
by
amix
18y ago
I have built a distributed key-value store on top of Tokyo Tyrant (which is a network interface on top of Tokyo Cabinet). I have also tried to use memcachedb (that's based on Berkley DB) and Tokyo Tyrant outperforms it - especially with lar
503.
▲
by
amix
18y ago
The major problem with electric cars is building an infrastructure for them. First, they can run around 100-150km (60 - 100miles) and after this they must recharge for a day or so. I.e. they are pretty useless on longer trips. In Denmark, D
504.
▲
by
amix
18y ago
Amazon Cloudfront forces you to put versioning if you want to expire objects manually, check out this page: http://docs.amazonwebservices.com/AmazonCloudFront/2008-06-3... (under `Object expiration`). Unless you specify "Cache-control: no
505.
▲
by
amix
18y ago
I have tried to search for MySQL vs. PostgreSQL benchmarks and most of them date to 2007. Are there any recent benchmarks that compare these databases? I personally use MySQL and I think it's an OK database (performance and installation wis
506.
▲
by
amix
18y ago
The problem with script.js?v={timestamp} is that it's ignored by some browsers while script_{timestamp}.js isn't. And with script.js?v={timestamp} you can't set good cache headers. Also, if you ever move to a CDN, then you are forced to use
507.
▲
by
amix
18y ago
From what I have read Facebook use a similar method: commit and deploy often and rollback if something messes up. We also use this method on Plurk.com and have done so for about a year. Thought, IMVU's case is pretty extreme :) The major pr
508.
▲
by
amix
18y ago
My parents have an African Grey parrot Benny and his intelligence is simply amazing. First, he can talk sentences and he mimics my mom pretty good. Second, he can connect sentences to meaning, so for example, when I enter the room he'll gre
509.
▲
by
amix
18y ago
It's kind of inspiring that they keep innovating or at least try to innovate with such an old product. I wonder how many engineers they have working on Gmail, does anybody know this?
510.
▲
by
amix
18y ago
We are different and a lot of "these" matters are subjective. E.g. some use Vim while others use Emacs and some people prefer to work from home while others prefer to work from an office. What's better is not that objective and people shoul
More ›