April 23rd, 2011
Sysbench-based benchmark comparing OLTP workloads for different builds and forks of MySQL on 64-core intel machine (Quad Intel Xeon L7555 ) . Only one definitive result is that MySQL cannot scale beyond 8 or 16 cores, or in some cases beyond just 2 cores (booo! ). In all cases performance went down when more than 24 cores were used .
Posted in benchmark, intel, Linux, mariadb, mysql, percona, performance, scalability | Comments Off | permalink
April 23rd, 2011
empirical data about difference in index sizes depending on the data type
Posted in data, dba, index, postgres, postgresql | Comments Off | permalink
April 23rd, 2011
Slides from Adrian Cockcorft, Netflix director of cloud systems about current state [for 2011] of Netflix's AWS-based architecture
Posted in amazon, architecture, aws, cloud, cloudcomputing, ec2, monitoring, netflix, simpledb | Comments Off | permalink
April 23rd, 2011
Notes from netflix tech blog about some key points of their EC2-based infrastructure design. Note that netflix is not affected by the current EBS problems in AWS US-EAST region. Their key point seems to be to equally split your infrastructure between 3 AZs (availability zones) and run at ~30% capacity, so if 2 AZs fail (as we have now) you would still running in one AZ @ 90%
Posted in amazon, architecture, aws, cloud, cloudcomputing, ec2, netflix, scalability | Comments Off | permalink
April 23rd, 2011
The subj. looks really bad!
Posted in amazon, aws, cloud, cloudcomputing, ec2, infrastructure | Comments Off | permalink
April 22nd, 2011
A blog post explaining nature of the current Amazon AWS outage. The deal is that several so-called availability zones (AZ) failed simultaneously in amazon US-EAST region, even though amazon's FAQ describe such event to be unlikely. Many interesting comments -- read them!
Posted in amazon, architecture, aws, cloud, cloudcomputing, ec2, hosting, outage | Comments Off | permalink
April 21st, 2011
"This announcement means that Dropbox never had any mechanism to prevent employees from accessing your files, and it means that Dropbox never had the crypto smarts to ensure the privacy of your files and never had the smarts to only decrypt the files for you. It turns out, they keep their keys on their servers, and anyone with clearance at Dropbox or anyone that manages to hack into their servers would be able to get access to your files. " -- ehh, a way to go dropbox :-( / noted at https://lwn.net/Articles/438401/
Posted in article, cloud, dropbox, encryption, privacy, Security, storage | Comments Off | permalink
April 21st, 2011
A review of three Lord of The Ring movies from someone who never read Tolkin's books before. Hllarios, read it of you have 30 minutes to spare. Taken from AVVA ( http://avva.livejournal.com/ )
Posted in culture, essay, funny, livejournal, lord_of_the_ring, movie, review, shortcut:moview, talkien, tolkien | Comments Off | permalink
April 19th, 2011
Facebook has build custom web-app testing framework to test server performance. Read it, especially the section on why "standard" SPECxxx benchmarks are not a good measure of anything. Also, ready about performance-per-watt, as this is really important for modern data centers.
Posted in architecture, benchmark, datacenter, facebook, intel, performance, server, spec, testing | Comments Off | permalink
April 19th, 2011
Notes about schema used by StackOverflow.com and related websites. It is very simple, only several tables used. (most likely because there are performance problems with joins)
Posted in architecture, database, databases, dba, design, mssql, schema, sql, stackoverflow | Comments Off | permalink
