Raise of minimum required BOINC Core Client version
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4312
Credit: 250417239
RAC: 35342
16 Nov 2010 12:44:57 UTC
Topic 195456
(moderation:
This is an early pre-warning: At some point in the next weeks or months, but definitely before the end of the upcoming GW run S5GC1HF we will raise the minimum BOINC Core Client version required for Einstein@home from 4.19 to 5.8.0. People using older Core Clients will need to upgrade to get work then.
BM
BM
Raise of minimum required BOINC Core Client version
)
You might want to put this on the front page and email it to all those using older versions. I know SOME people feel they cannot upgrade due to using their pc's at work, so you might lose some folks over this. Others will take the chance to do better Science! I would guess though that alot of us are already past that point. Good luck and I hope for the best!!
RE: You might want to put
)
I intended to do this when I have a bit clearer idea of the time we actually will do that.
Well, you should run BOINC "only on computers that you own, or for which you have obtained the owner's permission." What I'm more worried about are the people that run the stock client of their Linux distributions, some of which are pretty conservative. This includes some LSC computing clusters, btw.
The main feature that we need is the feature to compress the result file before it's sent back to the server. Currently the <10% of result files that are not compressed are taking up more than half of the storage space, I/O bandwith etc. on the server side. (We once used to compress the result files in the application, but BOINC dropped support for the boinc_zip library long ago and for good reason).
BM
BM
And here was me thinking that
)
And here was me thinking that my pair of win98se crunchers running 5.10.45 was an `old` version :)
The rest are on 6+.
RE: RE: You might want to
)
I saw you did it, it looks nice!
It totally agree but you know people!
Hmmm you learn something every day, I now know what I learned today!! Thanks!!
But why stop at version 5.8? Why not make people go all the way to 6.x instead? As stated by clive alot of us are already into the 6.x versions anyway. Versions 6.10.58 is the recommended version now and 6.12.6 is the latest development version available. I think 6.4.5 is the minimum to be able to use a gpu while crunching. And you are working on a new version of that software.
RE: But why stop at version
)
For all people manually upgrading their machines I strongly recommend to use one of the newest available Clients. However we do have some major contributors that still stick to a 5.8 client, and apparently upgrading is not easy for them.
BM
BM
RE: But why stop at version
)
BOINC 5.8.16 is the last version compatible with windows NT4!
Gruß,
Gundolf
Computer sind nicht alles im Leben. (Kleiner Scherz)
I have been unsuccessful
)
I have been unsuccessful getting BOINC v6.x to run on my older Macs as well. Not all of them are yet at v5.8 or higher, but I expect upgrades to that level will be feasible.
I've noticed a small problem
)
I've noticed a small problem with this when setting up host 3674581 for a burn-in test.
It's modern server hardware, so should perform well on Einstein tasks. But it's running Windows Small Business Server 2008 - which means it is running (has to run) as a Domain Controller. That in turn means that I'm restricted to BOINC v5.10.xx maximum. And that in turn means that all that lovely hardware is running base-level stock code, with no SIMD optimisations. Seems a shame, really.
RE: RE: But why stop at
)
WOW sooo many things to consider that I never had! Thanks for the info guys!
RE: But it's running
)
Check out this thread over at Collatz: http://boinc.thesonntags.com/collatz/forum_thread.php?id=455&nowrap=true#7137. Thanks to Crunch3r's method, I'm running 6.10.56 on a Windows 2003 Server domain controller under the Administrator account.