SPDY is new protocol proposed by Google as an alternative for HTTP(S). Currently Chrome and Firefox browsers are using it as default if available on server. It is faster in most cases by few to several percent. The side effect of using mod_spdy  external link is that it’s working well only with thread safe Apache’s modules. PHP module for Apache is not thread safe so we need to use PHP as CGI or FastCGI service. CGI is slow - so running mod_spdy for performance gain with CGI is simply pointless. FastCGI is better but it’s not possible to share APC  external link cache in FastCGI mode (ex. using spawn-fcgi), so it’s poor too. Best for PHP is PHP-FPM  external link which is FastCGI service with dynamic process manager and could use full advantages of APC. In such configuration I could switch from apache prefork to worker which should use less resources and be more predictable. ...

Yesterday I have problem with fglrx witch cause ugly system reset. After that, one of my drives was marked as failed in RAID5 array. Hotspare was automatically used to rebuild array. But this hotspare is the oldest and slowest drive I’ve got… After rebuild I’ve tested failed drive and it was fine - no bad block, no any other issue - so I wanted it running back in array. What I do: ...

I was configuring GlusterFS on few servers using Ansible  external link and have a need to update /etc/hosts with hostnames for easier configuration. I found this one working: - name: Update /etc/hosts lineinfile: dest=/etc/hosts regexp='.*{{item}}$' line='{{hostvars.{{item}}.ansible_default_ipv4.address}} {{item}}' state=present with_items: '{{groups.somegroup}}' Source: http://xmeblog.blogspot.com/2013/06/ansible-dynamicaly-update-etchosts.html  external link

Learn how to reset a user password in your own Ghost Blog by accessing the SQLite database and updating the password hash.

Learn about inodes exhaustion on XFS, how to check inode limits, and how to adjust the inode percentage to prevent issues on your Linux filesystem.

I’ve bought a NAS and customized it a little. But there was one thing which make my nights sleepless. NAS was seeking disks every 5~10 seconds - these was really irritating - especially when it was silent in room. I found that part of firmware was indexing or logging something so I wanted it dead! kill -9 was unsuccessful - process restarted after a while…. wrrr… I googled a little and found another signal I could use SIGSTOP, which will freeze process until I send SIGCONT to it - that was exactly what I need (because I normally use NFS/Samba and don’t need nothing more running on this device). ...

I’ve just bought new toy - Iomega StorCenter ix2-200 Cloud Edition. I have to play with few options before I could start using it. First thing - Firmware upgrade. Firmware upgrade I’ve started searching for latest firmware for ix2-200 Cloud and found that I have to register on Lenovo site to get firmware… I don’t like such sites where they force me to give all private data, but after few clicks on “Recommended articles” on that site I landed here: https://lenovo-eu-en.custhelp.com/app/answers/detail/a_id/26790  external link So it looks that I don’t need to register - point for me. ...

After some configuration changes I’ve stuck with VBP not listening nor on HTTP, nor on SSH port. Last resort was to use CLI to reenable HTTP access. Connect with parameters: Baud rate: 9600 Parity: none Bits: 8 Stopbits: 1 Flow control: none Then in login prompt you have to use login credentials (yes - they’re the same on every box (WTF?)): User: - root Pass: - @#$%^&*!() Password is shift + 2345678190 - there is 1 before 9! ...

I’ve crated this blog to get feedback from other IT guys about what I’m doing wrong (or not good enough). But this idea failed… I have only few comments on my blog (and about thousand and a half spams per month) - so, no feedback in my national language at all. Switching to English should make my audience bigger and I hope to have more attention thanks to that. This will be also a good practice of my English skill. I hope to find enough free time to translate some of older articles, but for now I’m thinking rather about changing engine of blog - to something more convenient. Maybe Octopress  external link or Ghost  external link . ...

Niedawno zainteresowałem się usługą Gearman i jedynej rzeczy której mi brakowało to jakiegoś łatwego mechanizmu zarządzającego workerami. Ale jak zwykle okazało się że inni mieli już ten problem i odpowiednie narzędzie istnieje - mowa o GearmanManagerze. Instalacja GearmanManagera Aby zainstalować GeramanManagera na serwerze gdzie już mamy Gearmana trzeba wykonać kilka kroków (wcześniej powinniśmy też zainstalować moduł gearmana do php’a): apt-get install git -y git clone https://github.com/brianlmoon/GearmanManager.git cd GearmanManager/install chmod +x install.sh ./install.sh Detecting linux distro as redhat- or debian-compatible Where is your php executable? (usually /usr/bin) /usr/bin Which PHP library to use, pecl/gearman or PEAR::Net_Gearman? 1) pecl 2) pear #? 1 Installing to /usr/local/share/gearman-manager Installing executable to /usr/local/bin/gearman-manager Installing configs to /etc/gearman-manager Installing init script to /etc/init.d/gearman-manager Install ok! Run /etc/init.d/gearman-manager to start and stop Worker scripts can be installed in /etc/gearman-manager/workers, configuration can be edited in /etc/gearman-manager/config.ini Mamy działającego GearmanManagera. Zalecam przyglądnięcie się plikowi config-advanced.ini bo jest tam kilka opcji, które warto dodatkowo ustawić. ...