Sadly though, the Linux machine's client did revert to same same http://www.worldcommunitygrid.org/forums/wcg/printpost_post,380983 and borged 7 tasks into nothingness. Worse, one [non-task] file got locked into partial download and blocked the https connect completion...

79 WCG 11-6-2012 21:17:09 [sched_op] Starting scheduler request
80 WCG 11-6-2012 21:17:09 Sending scheduler request: Requested by user.
81 WCG 11-6-2012 21:17:09 Reporting 11 completed tasks, requesting new tasks for CPU
82 WCG 11-6-2012 21:17:09 [sched_op] CPU work request: 87415.58 seconds; 0.00 devices
83 WCG 11-6-2012 21:17:42 Project communication failed: attempting access to reference site
84 WCG 11-6-2012 21:17:42 Scheduler request failed: Transferred a partial file
85 WCG 11-6-2012 21:17:42 [sched_op] Deferring communication for 1 hr 21 min 56 sec
86 WCG 11-6-2012 21:17:42 [sched_op] Reason: Scheduler request failed

only way out was to do a project reset, 4 good tasks [short CFSWs] that had completed taking with them. So much for yesterdays Linux adaptation.

edit: So, dnsmasq has not done the trick hoped for.

--//--

I don't recall seeing a report on Lubuntu 12.04 in this forum. I decided to try it on my old EEE PC yesterday (just from a thumb drive; haven't actually installed yet). So far it looks good, and feels much faster than the Mint 11 that I have on the same machine. I had no trouble installing BOINC through Synaptic, and it connected to WCG, downloaded a couple of WUs, and one of them ran just fine for as long as I had the machine on.

From the Lockergnome: Five Common Misconceptions About Linux

And yes, I have poorly running drivers for the NVidia card [above 280 nothing works proper]... hibernate [works], come out, and loads of jumbled screen elements and but pixel garbage on the screen. The only things that restore properly is the GNOME / Unity buttons and Dash/HUD. The only way to clear is log out, stop lightdm/gdm and restart that.

--//--

"Canonical has laid out their plans for handling UEFI SecureBoot on Ubuntu Linux. Similar to Red Hat paying Microsoft to get past UEFI restrictions, Canonical does have a private UEFI key. Beyond that they will also be switching from GRUB to the more liberal efilinux bootloader, and only require bootloader binaries be signed — and they want to setup their own signing infrastructure separate from Microsoft." http://linux.slashdot.org/story/12/06/22/1233...ting-past-uefi-secureboot Ubuntu's Plans To Implement UEFI SecureBoot: No GRUB2 http://www.phoronix.com/scan.php?page=news_item&px=MTEyNDY

Finally succumbed to the self afflicted pressure [signal 11 series in BOINC sending 12 hours of crunching south] and inserted several lines into cc_config.xml <options> section to suspend BOINC

<exclusive_app>apt-get</exclusive_app>
<exclusive_app>synaptic</exclusive_app>
<exclusive_app>synaptic-pkexec</exclusive_app>

This is in addition to having already got the BOINC suspend in place when non-BOINC system load is greater than 35% (just reduced to 30%). Somehow system updates are high enough, but lower still than this setting, but apps like DDDT2, CEP2 and HCMD2 just to sensitive... they cant stand it to not get near sole attention it seems.

How to do? On various Debian based distros [Ubuntu e.g.] navigate in a terminal window to /etc/boinc-client and do sudo gedit cc_config.xml to open the file [not the sym-link file in /var/lib/boinc-client]. Insert the lines and put the application mask names in place of ones given in the sample [as are showing in system monitor or top, case sensitive!]. Save the file and read it in with BOINC Manager or boinccmd [no client restart needed]. When successful, you'll be seeing these lines logged when launching your exclusive app, in my case synaptic:

28-Jun-2012 22:00:44 [---] Re-reading cc_config.xml
28-Jun-2012 22:00:44 [---] Unexpected text in cc_config.xml: ù
28-Jun-2012 22:00:44 [---] Unrecognized tag in cc_config.xml: <zero_debts>
28-Jun-2012 22:00:44 [---] Config: don't use coprocessors
28-Jun-2012 22:00:44 [---] Config: ignoring NVIDIA GPU 0
28-Jun-2012 22:00:44 [---] Config: don't compute while synaptic is running
28-Jun-2012 22:00:44 [---] Config: don't compute while synaptic-pkexec is running
28-Jun-2012 22:00:44 [---] Config: GUI RPC allowed from:
28-Jun-2012 22:00:56 [---] Suspending computation - an exclusive app is running
28-Jun-2012 22:00:56 [World Community Grid] [cpu_sched] Preempting dg05_c339_pla008_0 (left in memory)
28-Jun-2012 22:00:56 [World Community Grid] [cpu_sched] Preempting dg05_c331_pqa007_0 (left in memory)
28-Jun-2012 22:00:56 [World Community Grid] [cpu_sched] Preempting dg05_c331_pqb000_1 (left in memory)
28-Jun-2012 22:00:56 [World Community Grid] [cpu_sched] Preempting dg05_c331_pqa002_1 (left in memory)
28-Jun-2012 22:00:56 [---] Suspending network activity - an exclusive app is running
28-Jun-2012 22:02:40 [---] Resuming computation
28-Jun-2012 22:02:40 [World Community Grid] [cpu_sched] Resuming dg05_c339_pla008_0
28-Jun-2012 22:02:40 [World Community Grid] [cpu_sched] Resuming dg05_c331_pqa007_0
28-Jun-2012 22:02:40 [World Community Grid] [cpu_sched] Resuming dg05_c331_pqb000_1
28-Jun-2012 22:02:40 [World Community Grid] [cpu_sched] Resuming dg05_c331_pqa002_1
28-Jun-2012 22:02:40 [---] Resuming network activity

Long used by gamers who hate skipped screen frames, in my case, better loose 10 minutes on 4 cores than 12 hours of result errors, this time.
(with LAIM on, of course lossless... resuming where left off). Anyway, this will hopefully been the last time a too busy system causes tasks to crash. If there are more sporadically used apps causing this, they will be added as and when. :D

--//--

edit: added apt-get to the sample, for running package updates via terminal.