Improved phising filter and DKIM signing

I noticed an increasing number of phising emails from ‘Telekom’ and ‘DHL’ making it into user mailboxes. ClamAV anti-virus together with Sane Security Phising checks does a good job in stopping zero-day phising emails.

On a different account, DKIM Email signing seems to be preferred by Google these days. So time to add that too.

Let me know if you notice any problems with that.

OwnCloud Update

OC has been updated to version 5. Some of the shared files are not compatible between major versions. If you are missing any shares, just recreate them.

Rest should be ok. If you have any issues, let me know.

Happy New Year 2013!

Good luck in 2013 to everyone. I hope that it will be quieter than 2012 and we get some time to consolidate some of the big trends that started in 2012. My favourite ones are the Raspberry Pi and ownCloud.

I also applied the latest 3.5 update to WordPress. This brought some changes in media management. So in case you miss some pictures, just check for the correct paths in your posts. Generally most stuff should still work.

Discontinued: DaviCal

The outdated and little-used calDAV and cardDAV service at cal.snapdragon.cc:8443 has been discontinued in favor of the more modern ownCloud. Users are encouraged to move their data to the new platform. The OTA-profiles are already adapted to the new service. The service will probably be turned off on 1 Nov 2012 or after ownCloud gets better calendar sharing.

Filesystem Hierarchy Standard (FHS)

A crashed server at a friend’s reminded me of the importance of backups. Regular backups are in place of course, but scattered throughout /var/www, /media, /var/lib/mysql, etc. The FHS recommend to put everything in /srv/[protocol]. This should make backups a lot simpler because only /etc, /home and /srv need to be saved.

I changed the paths earlier today. Everything should still work fine and soft links for backwards-compatibility are in place for now. Please report any unexpected problems.

Hard Drive Failure

One of the drives in the RAID1-array failed for good last night. It was struggling for quite a while before that.

Some time in the next few hours, the server will be powered off to replace the drive and start RAID recovery.

Downtime will be around 20 min or less. I don’t expect any data loss and off site-backups are in place as well.

Network Maintenance in Denver

Due to network capacity increases there might be some downtime on Friday, Aug 27 from 23:00 CET (21:00 UPC).

We will be increasing our network’s transport capacity between Comfluent and our datacenter in Denver. We do not expect any downtime however there may be some routing changes during this maintenance window.

The maintenance is scheduled between 3PM-Midnight mountain time.

Thank you for your understanding.