Write heavy, high performance applications should probably use RAID 0 or avoid RAID altogether and consider using a larger n_val and cluster size. Read heavy applications have more options, and generally demand more fault tolerance with the added benefit of easier hardware replacement procedures.Good to see official guidance on this (via Bill de hOra)
(tags: via:dehora riak cluster fault-tolerance raid ops)
Facebook’s new erasure coding algorithm (via High Scalability).
Disk I/O and network traffic were reduced by half compared to RS codes. The LRC required 14% more storage than RS (ie. 60% of data size). Repair times were much lower thanks to the local repair codes. Much greater reliability thanks to fast repairs. Reduced network traffic makes them suitable for geographic distribution.
(tags: erasure-coding facebook redundancy repair algorithms papers via:highscalability data storage fault-tolerance)
Boundary’s Early Warnings alarm
Anomaly detection on network throughput metrics, alarming if throughputs on selected flows deviate by 1, 2, or 3 standard deviations from a historical baseline.
(tags: network-monitoring throughput boundary service-metrics alarming ops statistics)
My email to Irish Times Editor, sent 25th June
Daragh O’Brien noting 3 stories on 3 consecutive days voicing dangerously skewed misinformation about data protection and privacy law in Ireland:
There is a worrying pattern in these stories. The first two decry the Data Protection legislation (current and future) as being dangerous to children and damaging to the genealogy trade. The third sets up an industry “self-regulation” straw man and heralds it as progress (when it is decidedly not, serving only to further confuse consumers about their rights). If I was a cynical person I would find it hard not to draw the conclusion that the Irish Times, the “paper of record” has been stooged by organisations who are resistant to the defence of and validation of fundamental rights to privacy as enshrined in the Data Protection Acts and EU Treaties, and in the embryonic Data Protection Regulation. That these stories emerge hot on the heels of the pendulum swing towards privacy concerns that the NSA/Prism revelations have triggered is, I must assume, a co-incidence. It cannot be the case that the Irish Times blindly publishes press releases without conducting cursory fact checking on the stories contained therein? Three stories over three days is insufficient data to plot a definitive trend, but the emphasis is disconcerting. Is it the Irish Times’ editorial position that Data Protection legislation and the protection of fundamental rights is a bad thing and that industry self-regulation that operates in ignorance of legislation is the appropriate model for the future? It surely cannot be that press releases are regurgitated as balanced fact and news by the Irish Times without fact checking and verification? If I was to predict a “Data Protection killed my Puppy” type headline for tomorrow’s edition or another later this week would I be proved correct?
(tags: daragh-obrien irish-times iab bias advertising newspapers press-releases journalism data-protection privacy ireland)
_Bolt-On Causal Consistency_ [slides]
SIGMOD 2013 presentation from Peter Bailis, Ali Ghodsi, Joseph M. Hellerstein, Ion Stoica — adding consistency to an eventually-consistent store by tracking dependencies
(tags: eventual-consistency state cap-theorem storage peter-bailis)