In addition of object storage, Ceph is able to provide block devices. Ceph’s RADOS Block Device (RBD) provides access to block device images that are striped and replicated across the entire storage cluster. For example this is used to store Virtual Machines on OpenStack. With that solution, you’ve got a real fault tolerance system for your VM and distributed. The RBD part is an easy task when you’ve already setup a Ceph cluster.

Continue reading

Following the previous post, on Ceph Storage, I’ve updated my documentation with informations to play with OSD. I’ve covered : Setup OSD Remove OSD Use Ceph Ojects Storage Using the object storage system is not complicated and easily scalable. It’s a real pleasure to work with that kind of solutions. If you’re already working with other objects storage solutions, you won’t be lost.

Continue reading

I finally started to play with Ceph to add it in production. I’m still impressed by this storage solution. Just to remind, Ceph is an open-source, massively scalable, software-defined storage system which provides object, block and file system storage (not yet ready for production usage) in a single platform. It runs on commodity hardware-saving you costs, giving you flexibility. And as it’s in the Linux kernel, it’s easy to consume.

Continue reading

Cgroups permit to restrict a process to specific hardware like CPU or limiting IO disk etc… Managing it manually could be a nightmare and hopefully, a good workaround exists. It’s called cgconfig and allows you to have a running daemon to apply restrictions on all new running process. I’ve made a documentation for it and spent time on it as it’s not properly Debian packaged yet :-(

Continue reading

I recently needed to set a website in maintenance and wanted to do it as quickly as possible with any restart of Nginx. The idea is to add a maintenance.html page at the root of your web folder and then the maintenance page is automatically taken into account. To do it, simply follow the link (in french).

Continue reading

To be able to monitor the PHP-FPM service with connections, queue information, process… you’ll need to set the status page. This will give you something like this : pool: www process manager: dynamic start time: 18/Dec/2013:19:00:41 +0100 start since: 52972 accepted conn: 5268 listen queue: 0 max listen queue: 0 listen queue len: 0 idle processes: 3 active processes: 1 total processes: 4 max active processes: 4 max children reached: 0 Those informations are important to correctly tune the PHP-FPM service.

Continue reading