diff options
Diffstat (limited to 'content/posts')
| -rw-r--r-- | content/posts/git-on-rhel5.rst | 39 | ||||
| -rw-r--r-- | content/posts/piwik-in-docker-for-pelican.rst | 95 | ||||
| -rw-r--r-- | content/posts/recover-softraidlvm.rst | 2 | ||||
| -rw-r--r-- | content/posts/revive-euer-blog.rst | 92 | ||||
| -rw-r--r-- | content/posts/utf8-irssi-madness.rst | 15 |
5 files changed, 236 insertions, 7 deletions
diff --git a/content/posts/git-on-rhel5.rst b/content/posts/git-on-rhel5.rst new file mode 100644 index 0000000..d46229f --- /dev/null +++ b/content/posts/git-on-rhel5.rst @@ -0,0 +1,39 @@ +Install GIT on RHEL5 without Internet +##################################### +:date: 2014-02-18 13:37 +:tags: rhel, git + +I was facing the problem that i desperately needed git on one of the +development Redhat 5 systems which had no direct internet connection. As EPEL is an open +repository to retrieve all kinds of cool packages also Git is available there. +All i needed to do was to find out the dependencies, copy the packages to the +host and you are done. + +You need the following packages: + +- git (EPEL) + - perl-Git (dep) (EPEL) + - perl-error (EPEL) + - perl-termreadkey (EPEL) + - perl (installed in my case) (core) + +Today (2014-02-18) i downloaded the following packages: + +.. code-block:: bash + + #?/bin/sh + # on internet-available node: + wget 'http://mirror.bytemark.co.uk/fedora/epel/5Server/x86_64/perl-Error-0.17010-1.el5.noarch.rpm' \ + 'http://mirror.bytemark.co.uk/fedora/epel/5Server/x86_64/perl-TermReadKey-2.30-4.el5.x86_64.rpm' \ + 'http://mirror.bytemark.co.uk/fedora/epel/5Server/x86_64/perl-Git-1.8.2.1-1.el5.x86_64.rpm' \ + 'http://mirror.bytemark.co.uk/fedora/epel/5Server/x86_64/git-1.8.2.1-1.el5.x86_64.rpm' \ + # you may need this as well. + # http://mirror.bytemark.co.uk/centos/5.10/os/x86_64/CentOS/perl-5.8.8-41.el5.x86_64.rpm + scp *.rpm <lab-host>:~ + ssh <lab-host> + sudo rpm -i *.rmp + +If they go 404 have a look at +http://mirror.bytemark.co.uk/fedora/epel/5Server/x86_64/ for EPEL (replace +5Server with 6Server for RHEL6) and http://mirror.bytemark.co.uk/centos/ for +core packages for RHEL. diff --git a/content/posts/piwik-in-docker-for-pelican.rst b/content/posts/piwik-in-docker-for-pelican.rst new file mode 100644 index 0000000..1bf5df0 --- /dev/null +++ b/content/posts/piwik-in-docker-for-pelican.rst @@ -0,0 +1,95 @@ +Piwik for this blog +#################### +:date: 2014-02-18 08:00 +:tags: piwik, docker, pelican + +I was looking for a simple analytics tool for this blog. As everyone seems to +be using **piwik** and everyone loves **docker** for rapid deployment. As i am +always short on resources on all my machines i decided to deploy piwik on +another of my VMs which is ``mediengewitter.krebsco.de`` instead of directly on +``euer.krebsco.de`` Again it was a case of 'how-hard-can-it-be?'. +Turns out everything is much harder than expected. +As always this this is the digest of some hours work. + +Installing piwik in docker +========================== +After testing out all kinds of non-working Dockerfiles, this is what worked for +me: + +.. code-block:: bash + + git clone https://github.com/makefu/docker-piwik.git piwik + cd piwik + ./build + + # if you want persistence, see: + mkdir -p /media/ext/piwik/{www,mysql} + cat README + +Running piwik +============= +Because docker handles it's own network, the docker image port 80 must be +forwarded to the host (i use 10000). Also i want data persistence via exported filesystems. + +.. code-block:: bash + + docker run -p=10000:80 -d + \ -v /media/ext/piwik/www:/var/www + \ -v /media/ext/piwik/mysql:/var/lib/mysql piwik + # or ./run + +Adding piwik to pelican +======================= +Integration of piwik tracking to pelican **should** be straight forward, just +add the following to your ``pelicanconf.py``: + +.. code-block:: python + + PIWIK_URL='mediengewitter.krebsco.de:10000' + # first piwik site is always id 1 + PIWIK_SITE_ID=1 + +Bit nothing seemed to be happening, turns out my theme of choice does not +support piwik so i had to add this feature explicitly to the skin (by stealing +the code from another theme). I created a pull request for my code: +https://github.com/getpelican/pelican-themes/pull/195 . +If it will never be merged, use my repository for themes: + +.. code-block:: bash + + cd <pelican-dir> + git clone git@github.com:makefu/pelican-themes.git -b add-piwik-to-gum themes + +Add piwik to another skin +========================= +If you are using another skin without piwik integration, this is basically what you +need to do: + +1. Add piwik.html to '``<skin>/templates``' + +.. code-block:: html + + {% if PIWIK_URL and PIWIK_SITE_ID %} + <script type="text/javascript"> + {% if PIWIK_SSL_URL %} + var pkBaseURL = (("https:" == document.location.protocol) ? "https://{{ PIWIK_SSL_URL }}/" : "http://{{ PIWIK_URL }}/"); + {% else %} + var pkBaseURL = (("https:" == document.location.protocol) ? "https://{{ PIWIK_URL }}/" : "http://{{ PIWIK_URL }}/"); + {% endif %} + document.write(unescape("%3Cscript src='" + pkBaseURL + "piwik.js' type='text/javascript'%3E%3C/script%3E")); + </script><script type="text/javascript"> + try { + var piwikTracker = Piwik.getTracker(pkBaseURL + "piwik.php", {{ PIWIK_SITE_ID }}); + piwikTracker.trackPageView(); + piwikTracker.enableLinkTracking(); + } catch( err ) {} + </script><noscript><p><img src="http://{{ PIWIK_URL }}/piwik.php?idsite={{ PIWIK_SITE_ID }}" style="border:0" alt="" /></p></noscript> + {% endif %} + +2. Add this line somewhere near the end but before </body></html> to '``<skin>/templates/base.html``': + +.. code-block:: html + + ... + {% include 'piwik.html' %} + ... diff --git a/content/posts/recover-softraidlvm.rst b/content/posts/recover-softraidlvm.rst index efc070d..a2cd31a 100644 --- a/content/posts/recover-softraidlvm.rst +++ b/content/posts/recover-softraidlvm.rst @@ -1,6 +1,6 @@ Recover Softraid/LVM #################### -Date: 2012-02-06 10:24 +:date: 2012-02-06 10:24 :tags: mdadm, softraid, lvm MD Array fails to assemble diff --git a/content/posts/revive-euer-blog.rst b/content/posts/revive-euer-blog.rst new file mode 100644 index 0000000..a6470ce --- /dev/null +++ b/content/posts/revive-euer-blog.rst @@ -0,0 +1,92 @@ +Revive this Blog +################################## +:date: 2013-02-17 14:26 +:tags: openssh, dropbear + +2 years ago i lost this blog in an unexpected VPS shutdown and of course i had +no backups (duh!) and had not used any kind of version management. +Thanks to archive.org at least the content of the blog left intact. I guess +that's why it is sometimes called **'The WaybackUp Machine'**. + +My old blog can be found at +https://web.archive.org/web/20121213091551/http://euer.krebsco.de/ and +even the rss feed is intact. + + +The old blog was created using **octopress**, a static site generator written in +ruby. Because the markdown files were essentially lost i thought it is time to +try out something new. +I chose **pelican**, a static site generator written in python because in a +worst-case scenario i could fix the python code. I will also test out +reStructured Text instead of Markdown. + +Installation of pelican +----------------------- + +.. code-block:: bash + + #?/bin/sh + virtualenv my-blog + cd my-blog + . bin/activate + pip install pelican + pelican-quickstart + ... + +Importing RSS +------------- +Pelican supports importing old rss feeds. +I tried importing the archive.org rss feed but besides the date and title of +the post and the date the markup was pretty much broken, the code with line +numbering resulted in a broken <pre>-table. This happens for rst and markdown +output. + +In addition the importer uses pandoc, a haskell markup transformer which is +with all dependencies like 100mb in size. + +For the records here is what was needed: + +.. code-block:: bash + + #? /bin/sh + . bin/activate + yaourt -Sy aur/pandoc-static + pelican-import --feed https://web.archive.org/web/20120709004415/http://euer.krebsco.de/atom.xml -o content/posts + # cleanup all the posts in content/posts + +Configure pelican +----------------- +Pelican needs to be configured in ``pelicanconf.py``: + +.. code-block:: python + + AUTHOR = 'makefu' + SITENAME = 'only code is pure' + # rss feed to be built + FEED_ALL_ATOM = 'feeds/all.atom.xml' + # for RSS in the headline + MENUITEMS = (( 'RSS', '/feeds/all.atom.xml'),) + + # add robots.txt + STATIC_PATHS = [ 'extra/robots.txt', ] + EXTRA_PATH_METADATA = { 'extra/robots.txt': {'path': 'robots.txt'}, } + + # twitter link + SOCIAL = (('@makefoo', 'http://twitter.com/makefoo') ,) + # add disqus comments + DISQUS_SITENAME = 'euer' + # all the other lines of config + + +Configure themes +---------------- +Themes need to be retrieved separately. + +.. code-block:: bash + + #? /bin/sh + # omit --recursive for a subset, gum is in the core + git clone --recursive https://github.com/getpelican/pelican-themes themes + + # add the theme line to pelicanconf + echo 'THEME="themes/gum"' >> pelicanconf.py diff --git a/content/posts/utf8-irssi-madness.rst b/content/posts/utf8-irssi-madness.rst index 5b6fb87..fb86115 100644 --- a/content/posts/utf8-irssi-madness.rst +++ b/content/posts/utf8-irssi-madness.rst @@ -1,25 +1,27 @@ -Title: Utf8 in an Irssi/tmux/putty/windows Stack -Date: 2012-06-22 -Tags: irssi, utf8 +Utf8 in an Irssi/tmux/putty/windows Stack +######################################### +:date: 2012-06-22 00:00 +:tags: irssi, utf8 Getting irssi running with utf8 support in a putty/tmux stack is madness. Here is what you have to do. add lines in .{ba,z}shrc: - :::bash +.. code-block:: bash + export LANG=en_US.utf8 export LC_ALL=en_US.utf8 add lines in .tmux.conf: +.. code-block:: bash - :::bash set-option -g default-terminal "rxvt" set-window-option -g utf8 on in irssi: +.. code-block:: bash - :::bash /set term_charset UTF-8 /set recode_autodetect_utf8 ON /set recode_fallback UTF-8 @@ -30,6 +32,7 @@ in irssi: /quit in putty config: +.. code-block:: bash window -> translation -> Received data assumed to be in which character set: UTF-8 -> Use Unicode line drawing code points |