Monthly Archives: April 2016
The second level LPIC certification (LPIC-2)

As I mentioned in the first text of the certification series, I went further upstream and got myself the second level of what LPI has on offer – LPIC-2 (verify). Was it easy? Was it hard? Let’s start.

Introduction

The LPIC-2 certification is granted by passing 2 exams, but only if one already holds a LPIC-1 certification. This means that in order to get LPIC-2 certified, one must pass a total of 4 exams.

  • Exam 201 – with a focus on advanced system administration topics such as kernel and boot loader configuration, filesystems and troubleshooting.

  • Exam 202 – focusing on a couple of common services such as e-mail, http, proxies and file sharing.

As with LPIC-1, the exams can be passed in any order, certification being granted when passing both.

Logistics

I have passed both exams with a Pearson VUE test center (actually 2 of them, as I have gone to 2 different locations for the exams). In this setup each exam normally costs around EUR 150; the second was paid by my current employer.

Preparation

What I have mentioned, preparation-wise, in the first text is still valid. Books alone will not help pass such exam, neither a light hands-on experience. The questions themselves are by no means tricky – if one actually used that particular piece of software or went through that usage scenario, then the answer comes fast; otherwise it won’t and the coin toss won’t help either. What I have also noted is that some questions are (randomly) asked at both 1st and the 2nd level; it’s very likely that the border between “harder” 1st level topics and “easier” 2nd level ones is very shallow.

Continue Reading →

Variable-Size Directories

From time to time one may receive a request from the QA in line of:

For testing purposes, I need that /opt/test/xxx directory be limited to 10 Megabytes. This directory is used by the zzz application ran as user tester.

How could the directory size be limited in Linux? Is it even possible? – these are fair questions and the answer is yes. One needs to:

  1. Use the directory as a mount point for a size-limited storage device;

  2. Use the proper mount options to allow full access to the non-root user specified;

  3. Disable Selinux (easy) or allow that particular user to access data on mount points (complicated).

Let’s start with the beginning, the storage device. There are multiple options here:

  • A simple loop device (a regular file used as a file system);

  • A logical volume (LVM), assuming the disk setup is based on this technology and there is enough free space left to accomodate the new device;

  • Attaching a new storage device (e.g. in a Cloud environment like Amazon Web Services).

Continue Reading →