Back in 2009/10 when I had planned to set up a little SETI server farm, and was looking at different ways of doing it, I came across Dotsch's (Lars) software. He had released 3 different versions of his program, the last in 2010, v1.2.

The last time he posted on here was at the end of 2010, and the last update he did to his software was in 2013, after a 2 year absence from anything new on his site, to get version 7 working, and after that he sort of fell off the face of the earth.

I emailed him at both addresses on his site, the SETI one bounced back, but the other didn't. Hopefully he'll answer, as his stuff was I thought pretty cool, and would like to give it a shot, but sadly in it's current state, it won't work to crunch, as it is too old.

If anyone has any info, or another contact for him, or even knows if he has given up on any more development of his software and retired from SETI software work, I'd love to hear about it.

---

Don't know other than his computers still connect on occasion but he has a very low RAC. http://setiathome.berkeley.edu/show_user.php?userid=179343

---

I remember helping him seed his "software" via torrents back in the day. It was literally just Ubuntu live-boot CDs that already had BOINC and Lunatics installed. Really. That's all it was.

---

Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)

Hmm, I seem to recall there was more to it than that, and I think that at least one of the purposes/goals of it was to allow Linux non-natives to install a multiple system farm "fairly easily". Just before I backed off here, I had tried starting the config of I believe 2-3 systems, booting off the network and not having to have hard drives on them.

It was actually fairly close to being prime time, and I was just bad enough at it that I ended up helping him find some issues that weren't obvious to someone who knew what they were doing. But, add in someone fairly clueless who didn't know the proper procedures (yours truly), well, lets say I was an ok, although unwitting beta tester. ;-)

His site is still up, http://www.dotsch.de/boinc/Dotsch_UX.html and lists the install instructions for what I was doing. I presume someone could take whatever the current Linux flavor (that is most favorable for running BOINC), and cobbling together/configuring a system that can do what he was working on, but you would have to know exactly what you were doing.

Why bother at all, when Windows is out there and is so easy? Well, for one thing I would think that running Linux would be less resource hungry than Windows, all things considered, over the long run, you would get more production out of Linux than an identical Windows setup.

Plus the ability to just boot off the network, using a KVM switch, and not having to have the cost, heat and energy useage of the hard drives would also be a bonus. Granted, small modern 2.5" HDD's don't cost _that_ much or use as much energy as the ones from years back, and an SDD is even less, though more $.

USB is also an option, but I would think in a "set and mostly forget" (that is until version 9 or whatever comes out) type of setup, running it off a network would be the most efficient. Your thoughts on any or all of it? I am always keen on learning new things and hearing different perspectives.

---

In my job as a software test engineer I've had access to many types of hardware and very OS I wanted to use. For CPU applications I never found one OS to be measurably better for running BOINC projects than any other.
My preferred OS that I used in my lab was actually Windows 7 64-bit. Simply because it made managing 30+ machines easier for me. All of my test machines used ghost images for the OS. So when I was done doing testing I would load up a clean copy of Windows 7 & start BOINC. Given most of the machines had 2-4GB of RAM memory usage wasn't a problem.

I also toyed around with BOINC PE Which is a preconfigured version of Windows PE with BOINC. However some of the limits of Windows PE made it less desirable vs using a full version of Windows for me.


For a farm of diskless SETI@home is pretty light on disk activity. With a limit of 100 tasks the most a host would need to worry about is ~800MB if it had 100 AP tasks or ~40MB for 100 MB tasks. So a 120GB SSD would probably serve as plenty of space for quite a few diskless systems.

The base image of Dotsch UX could be updated with the current release of BOINC and optimized CPU apps. Then deployed exactly as the instructions indicate.

---

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[

Yeah, I am most familiar with windows as well, running either XP or 7 for the most part. I don't have the expertise to even fathom where to begin with bringing what he had created up to date, though I know there are a number of Linux Wiz's around here... ;-)

---

What do you mean it doesn't work?
It works for me.

---

Mike Bader
BOINC V7.16.5
http://setiathome.berkeley.edu/team_join_form.php?id=5 - Join Our International Team
[img]http://boinc.mundayweb.com/one/stats.php?

Well, I guess I shouldn't make the assumption that it doesn't, but I saw how my 2 Windows systems at the inlaws stopped working after v7 tasks were done, I guess I just assumed that the same thing would happen for this one.

Also, I think he was running on v10 or 11 of Ubuntu, not sure if that is a good or bad thing, but thought that if he was still into it full steam like in the past, he might be using a little more current version of Linux if for no other reason than bug/security patches?

I am very much a featherweight in terms of knowledge on any flavor of Linux, that's why I'm posting here, to rub shoulders with those who are in the know, and might have better information than I do.

---

I'd be interested if you find a new version.
Or someone else takes up the task.
Pretty sure I am getting SETI v8 tasks that are completing and validating.
My computers are sitting in a corner headless most of the time and are older.
Yeah the linux version is old. But running nothing else and from a CD
so I think it is pretty safe.

http://setiathome.berkeley.edu/host_app_versions.php?hostid=7834218

http://setiathome.berkeley.edu/show_host_detail.php?hostid=7834218

regards

---

Mike Bader
BOINC V7.16.5
http://setiathome.berkeley.edu/team_join_form.php?id=5 - Join Our International Team
[img]http://boinc.mundayweb.com/one/stats.php?

At Al's behest, will probably also be looking into the situation as the Cuda v8 application situation eases.

Can't make any promises, as will probably be tying up Cuda multibeam cross platform for a while yet, but will probably put a headless configuration onto some roadmap, examine what was done before, and issue a call for help.

My focus has been applications, and tooling up to gradually move applications under a unified buildsystem. Probably though, the advent of GPU passthrough into VMs, and fast home networking, will warrant headless crunching farms again, along with custom clients.

---

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.

Thanks, Jason! You're a Rockstar.

---