The “Alien Pastures” blog has been moved into a Virtual Machine (hence the short downtime yesterday) after its previous host hardware was replaced with a much beefier machine. This was a nice opportunity to configure the VM with a SSL certificate issued by LetsEncrypt (again, thanks Robby Workman).
So, from now on, the blog is served with a HTTPS URL by default. I hope you understand, this is a process seen on many other web sites too.
My “Alien Pastures” blog will be down for two or three hours on Thursday, 27 July 2017, starting around 1700 UTC.
The blog runs on a server in the Teklinks datacenter, and that will be replaced by new hardware. Since this involves driving to the datacenter and doing physical work (thanks Robby Workman!) the slightly longer-than-usual downtime is needed.
Additionally, the following services that are running on the same hardware, will also be down:
For the sixth time in just 5 days I had do a system_reset on my virtual machine which runs “taper.alienbase.nl” as well as “docs.slackware.com“. The virtual machine is crashing under the load that is put on it by demanding rsync processes. According to my pal who donated the use of this VM to me for free, the rsync download rate is at a continuous 100 Mbit/sec for most of the time. This is apparently too much for the server, as well as for my pal who had not anticipated this kind of bandwidth consumption. He has been paying quite a bit of extra money for the excess bandwidth during the past months.
I do not like it when someone loses money because of me, so drastic measures are required.
As of tomorrow (monday 15 Feb 2016) I will kill the rsync access to taper.alienbase.nl.
HTTP access will still be allowed, so people who configured slackpkg to download from my server will still be served.
I hope that this will put my pal at ease again, and also prevent the server from crashing.
Apologies for the inconvenience this causes to people who use rsync to keep their local mirrors up to date. I will try to find another location to host my SlackBuild repository as well as the multilib and KDE (ktown) repositories. The downtime of this VM caused downtime for docs.slackware.com as well which I think is unacceptible.
Two years ago (on 29 february 2012), the Raspberry Pi Model B went on sale. More than 2.5 million Raspberry Pis have been sold to date! An amazing number, considering that the original goal was to equip british school kids with cheap hardware for Computer Science education.
Thanks to these enormous sales numbers, the Raspberry Pi Foundation (a not-for-profit organisation) was able to sponsor several Open Source projects writing code which can be used with the hardware (XBMC, libav and many others).
And now, two years later, there is a new surprise. The Raspberry Pi has been developed as “open” as possible, however there was a part of the hardware which was not open: the VideoCore IV 3d graphics core on the Broadcom application processor for which only a “binary blob” exists and which is addressed by a thin layer of Open Source graphics kerneldriver. This is not unusual – most if not all of today’s ARM-based mobile hardware has a closed-source graphics stack and no public register-level documentation of the hardware.
This is changing now! As announced on their blog, Broadcom has decided to open up their VideoCore IV 3d core to accompany the two-year anniversary of the Raspberry Pi. The code of the graphics stack has been open-sourced under a liberal 3-clause BSD license and it’s accompanied by complete register-level documentation for the graphics engine. This is unique for the ARM hardware platform as far as I know.
If you are an experienced hacker/programmer, you may be up to the challenge posed by the Raspberry Pi Foundation: to port the open-sourced graphics stack (for the BCM21553) to the Raspberry Pi’s processor (BCM2835). And they will pay you a bounty of $10,000 if you are the first person to demonstrate satisfactorily that you can successfully run Quake III at a playable framerate on Raspberry Pi using your ported drivers.
How cool is that? Of course I hope it will be a Slackware hacker who will reap this reward.
According to the story told by that server’s owner, the power cord got loose during the re-build of the RAID array after the guy replaced a failing hard drive.
After a power-on the RAID volume refused to mount…
It looks like the host server may be up and running at the end of the day, if the owner manages to recover from the partition corruption. In that case, my mirror server taper.alienbase.nl and the Slackware Documentatoin Project’s Wiki docs.slackware.com may be online a little later, provided that I can access the server and start the Virtual Machine.
In the worst case, the server will not be able to recover and I have to find a location where I can restore my daily backup of the Wiki content. It’s just bad co-incidence that I will travel to India tomorrow early morning. I have no idea how fast I would be able to arrange for a new host, fix the DNS entries, setup apache and start the Wiki… for the docs.slackware.com hostname I need to involve Pat.
More news later.
Update Fri Nov 23 18:55:29 UTC 2012 –
The server is back up! The server admin fixed it, no data was lost.
Dear visitor, you seem to be using an Ad Blocker. Please consider whitelisting 'Alien Pastures'. I use the revenue from displaying ads (small as it is) to keep this site running. Thanks!