397 liens privés
Un tuto très simple qui explique comment utiliser l'excellent pgbadger.
C'est un script perl qui analyse les logs postgres, hors connexion.
Il génère une page html (avec tout le contenu statique nécessaire "inliné") qui présente tout un tas de statistiques dont les fameux histogrammes des requêtes lentes, etc..
Plus d'informations sur l'outil: http://dalibo.github.io/pgbadger/ + https://github.com/dalibo/pgbadger/
Encore un bel outil postgres, signé Dalibo: http://www.dalibo.com/ <3
There’s been a welcome focus in the Docker community recently around image size. Smaller image sizes are being championed by Docker and by the community. When many images clock in at multi-100 MB a...
En gros ils rappelent que chaque instruction d'un Dockerfile donne lieu à un commit. Ils expliquent donc comment éviter des commits qui font que les images résultantes sont inutilement volumineuses.
"""
In case the number of tomcat threads and acceptCount values are set to be too high, a sudden increase in traffic will fill up the OS queues and make all the worker threads busy. When more requests than that can be handled by the system are sent to the machines, this "queuing" of requests is inevitable and will lead to increased busy threads, causing cpu starvation eventually. Hence, the crux of the solution is to avoid too much queuing of requests at multiple points (OS and tomcat threads) and fail fast (return http status 503) as soon the application's maximum capacity is reached. Here is a recommendation for doing this in practice:
Fail fast in case the system capacity for a machine is hit.
"""
Ernie Souhrada, DBA chez Pinterest, explique comment il a drastiquement amélioré les perfs mysql sans changer de hardware.
"""
When we enable all of the optimizations, we find we can achieve roughly 500 percent more read and write throughput at both 16 and 32 threads while simultaneously reducing p99 latency by over 500ms in both directions. On the read side, we go from approximately 4100 – 4600 QPS to just over 22000 – 25000, depending on concurrency. On the write side, we go from approximately 1000 QPS to 5100 – 6000 QPS. These are massive gains in headroom and performance achieved with just a few simple changes.
"""
sur du i2.4xlarge (de la grosse instance ec2 de porc), percona-server-5.6, ubuntu 12.04.
Voir aussi http://www.slideshare.net/denshikarasu/all-your-iops-are-belong-to-us-a-pinteresting-case-study-in-mysql-performance-optimization
qui va plus dans les détails, avec les tests sysbench pour les différents disques, versions de kernel, modifs my.cnf, irqbalance, options de montage, etc... qui ont permi de tirer les conclusions.
On apprend notamment que le kernel en version 3.18 montre de meilleurs perfs IO qu'en 3.13.
Très pointu.
certains tirent profits des ssd ephemeres fournis avec les c3 pour les mettre en tant que cache devant des ebs.
Ce service est une tuerie.
En plus le code est libre https://github.com/WPO-Foundation/webpagetest
Pour faire tourner sa propre instance : https://sites.google.com/a/webpagetest.org/docs/private-instances
un bon guide de démarrage pour optimiser un site web
Bootstrap, as of version 2, requires jQuery for some components such as dropdowns, alerts and the little "hamburger" icon that triggers a collapsible menu. While jQuery is an excellent cross-browser framework, it can be overkill if it's only used for displaying and hiding things. Of course, if a project already uses jQuery then there's no problem, but if not here's a script to replace jQuery and some of the Bootstrap plugin scripts. Using this should reduce download size and speed up loading time for your users.
code : https://github.com/tagawa/bootstrap-without-jquery
J'ai allégé mon site (fralef.me) de 119.3K en supprimant jquery (1.7.2) et en mettant ce fork de bootstrap.min.js. joie!