LHC@home

It has been another two weeks and I'm sorry about the silence. Sadly we've had some hiccups with the systems we work on that are not BOINC (as I've explained we are not on BOINC full time, not even close) and that has meant this project was out on the backburner and I apologise for that. Also it's the start of conference season so between attending and organising things where you learn a lot but get nothing done it has slipped again.

I will get a fuller update to you when I can but at least some WUs went up recently.

Thanks for the update Neasan.

---

What about people who have attached, zero credit, zero rac for the simple fact they couldn`t get any work.

What about them?

---

What about people who have attached, zero credit, zero rac for the simple fact they couldn`t get any work.

What about them?

The account would still be active, it's just the computers that would be deleted. The next time the computer requests work, it would be added to the database. No different than if / when a existing user adds a new computer to his account.

---

It has been another two weeks and I'm sorry about the silence. Sadly we've had some hiccups with the systems we work on that are not BOINC (as I've explained we are not on BOINC full time, not even close) and that has meant this project was out on the backburner and I apologise for that. Also it's the start of conference season so between attending and organising things where you learn a lot but get nothing done it has slipped again.

I will get a fuller update to you when I can but at least some WUs went up recently.

I live for the every two week update. :) Anything more to report this time around? (With fingers crossed.)

What about people who have attached, zero credit, zero rac for the simple fact they couldn`t get any work.

They wait, just like the rest of us?
It shouldn't cause any problems when the project does finally get work. Even if the host were deleted, it should be able to reattach itself using the info in the file on the host itself ('s always worked for me).

---

sQuonk
Plague of Mice
Intel Core i3-9100 CPU@3.60 GHz, but it's doing its bit just the same.

8-) dont worry, the pace the work is progressing every one will some day or another detach themselves B-)Neasan will be left talking to himself ;-)then when the work finally arrives he will be the only one left crunching them ;-) LoL
Regards
Masud.

---

I've explained we are not on BOINC full time, not even close

I don't wish to seem rude, but if that is the case, why was reponsibility for LHC@Home transferred to QMC?

---

8-) dont worry, the pace the work is progressing every one will some day or another detach themselves B-)Neasan will be left talking to himself ;-)then when the work finally arrives he will be the only one left crunching them ;-) LoL

I've explained we are not on BOINC full time, not even close
I don't wish to seem rude, but if that is the case, why was reponsibility for LHC@Home transferred to QMC?

I want to second these statements.

Technically, I'm only still attached to LHC by the record, but have everything suspended. While other Projects (even Beta ones) quickly kick into some action and get work going, LHC just sits there and silently accumulates Long Term Debt which doesn't help me any.

Since it's been well over 3 months now (again, it's not like we've been waiting for almost a year), I would have wished for the Admins to at least temporarily shift priorities to the BOINC setup to get it started.
IMHO, having it delayed until it's all shiny and runs 100% bug-free from day 1 (which it won't anyway ;) ) is nice but with no Userbase remaining...of little value.

It'll be interesting to find out how many active Crunchers the Project will still have when it goes back alive... Some day in the future...

---

Scientific Network : 45000 MHz - 77824 MB - 1970 GB

I still find this to be the project that best proves the BOINC system is a success. There is no way a stand alone project could survive with with the intermittant supply of work that this project has had since it's launch. With BOINC it is no big deal it just sits idle on hosts until work is available then it goes into action.

Comparing this project to a beta is inappropriate. This project has never had a steady supply of work. There was a period when it was starting that the studies were coming fast enough that the server would rarely run dry, but this is still a production level project with intermittant work.

---

BOINC WIKI

BOINCing since 2002/12/8

Agreed on the flow of work, but a constant flow (after seeing the last ~2 years) is not expected by anyone anymore.

Having it deliver no work, however, is only "no sweat" for BOINC Installations attached to LHC only.
Add other running Projects, and you'll find that their LTD is growing into the Minus, while the month-long idle LHC will amass Millions of Seconds LTD, sending all remaining active Clients deep into "Overworked" Status and even the latest BOINC Versions (expecting LHC to deliver anytime) eventually freaking out and going into "x Deadline misses" mode.

So, no, it's no big deal if you know how to handle it, but it becomes when you let it run along and let it completely screw up the work scheduling over time.
BOINC was designed for temporary outages / work shortages of several Projects at once. But not for a Project that technically exists but actually just sits there with no or insufficient amounts of work for a year.

Looking at the minimal amount of work distributed over the last 12 month, BOINC LHC was actually not needed. A medium sized farm right in the Lab would have had more than enough time to complete it... all by itself.

IMHO, Projects that just about never have work can hardly be called a success for BOINC. When it goes back into production (at whatever levels), we will see how many crunchers it still has...
Good news is, at least we are still there ;)

---

Scientific Network : 45000 MHz - 77824 MB - 1970 GB

I've had LHC permanently active along with half a dozen or so other projects for the last few years and I have never had any scheduling problems. The only long-term concession I have made to the intermittent work from LHC is to make its priority about 10 times that of the other projects. If the work becomes more regular I'll reduce it again.

---

Projects that are out of work, no new work, or suspended do not build LTD like a normal project. Once the project's LTD gets to the point where it is the first project contacted (has the most positive LTD) it's LTD will barely move provided another project does not cause EDF mode. It does not stop completely, it will gain a few seconds due to the delay between the expiring of the defferal and the RPC response with the next defferal. Rebalancing the LTDs will also cause some creeping.

On this host LHC, pirates, and renderfarm have some serious LTD, ~100k secs, however none of the normal projects have more than one day of negative LTD and they get work pretty consistantly.

---

BOINC WIKI

BOINCing since 2002/12/8

I still find this to be the project that best proves the BOINC system is a success. There is no way a stand alone project could survive with with the intermittent supply of work that this project has had since its launch. With BOINC it is no big deal it just sits idle on hosts until work is available then it goes into action.

Hear, hear!

---

sQuonk
Plague of Mice
Intel Core i3-9100 CPU@3.60 GHz, but it's doing its bit just the same.

@Keck Komputers

Now that's odd (because I've heard several folks saying that over a longer timespan).

Whatever BOINC Version I had installed at any given time, LHC's LTD was far exceeding 2 Million, crippling active Projects on V5.4x (seeing even 5 day cache sizes being reduced downto as little as 30 Minutes per Host), even on V5.9.3 I just had to begin suspending and resetting LHC to remove LTD's exceeding 1 Million and again crippling active Projects...

Had I not manually intervened, I would have eventually seen the entire Network run dry, as active Projects reverted to "fetching no new work" as BOINC was literally starving to get work off LHC and satisfy its massive accumulated LTD.

Not sure why, but - because I've heard exactly opposite statements like yours - apparently not all are affected by this (?)
I'm basically left with no other choice short of detaching in order to keep my Network just running. All this regardless of keeping everything offline most of the time (feeding from refreshing caches once or twice a day) or keeping everything online 24/7.

*scratches head*

---

Scientific Network : 45000 MHz - 77824 MB - 1970 GB

It could be because I stay on the latest development version for testing, so I've been running something better than 5.4.x for a long time. The 5.4.x versions would get pretty ugly, but the 5.8.11+ versions work much better, it may take some time to repair the damage from earlier versions though.

---

BOINC WIKI

BOINCing since 2002/12/8

It's boinc debt view. For link, ask google.

---

Yep, tried that and found it wasn't Network capable at that time.

Anyway, the Problem is solved again and my 5.8.15, 5.8.16 and 5.9.3 Installations are in the process of recovering back to normal.

---

Scientific Network : 45000 MHz - 77824 MB - 1970 GB

It has been another two weeks and I'm sorry about the silence. Sadly we've had some hiccups with the systems we work on that are not BOINC (as I've explained we are not on BOINC full time, not even close) and that has meant this project was out on the backburner and I apologise for that. Also it's the start of conference season so between attending and organising things where you learn a lot but get nothing done it has slipped again.

I will get a fuller update to you when I can but at least some WUs went up recently.

Three weeks...Bump for an update?