Front page says bucket loads of work available, but for the last couple of hours I've been getting...

12/12/05 15:56:56|rosetta@home|Sending scheduler request to https://boinc.bakerlab.org/rosetta_cgi/cgi
12/12/05 15:56:56|rosetta@home|Reason: To fetch work
12/12/05 15:56:56|rosetta@home|Requesting 8640 seconds of new work
12/12/05 15:56:56||request_reschedule_cpus: files downloaded
12/12/05 15:57:01|rosetta@home|Scheduler request to https://boinc.bakerlab.org/rosetta_cgi/cgi succeeded
12/12/05 15:57:01|rosetta@home|No work from project
12/12/05 15:57:06|rosetta@home|Deferring communication with project for 3 minutes and 57 seconds

*** EDIT ***

I now nottice the other thread about server issues, which was not on this subject, has moved onto this subject!

---

Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.

My evesham machine was dry for a few hours, (Rosetta anyway - had Predictor wu's coming and going), but it has grabbed a wu now.

---

Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.

Copied from Homepage Technical News -

December 12, 2005
Our work unit feeder is having a tough time keeping up with all the client requests for work. A short term fix (as has been done before), is to optimize the database tables. We will be doing this later today at 3pm and also backing up the database. As stated before, we are going to expand our servers soon to deal with this issue.

---

Copied from Homepage Technical News -

December 12, 2005
Our work unit feeder is having a tough time keeping up with all the client requests for work. A short term fix (as has been done before), is to optimize the database tables. We will be doing this later today at 3pm and also backing up the database. As stated before, we are going to expand our servers soon to deal with this issue.

Yeah, I just installed BOINC 5.3.2 for the red lines (and some other things)


Click the pic!

and I really thought that Rosetta had problems running in that, as there wasn't any Rosetta WU's downloading, but then I read this message! Phew!

So let's see how Rosetta deals with the 5.3.2 when it's up again!

---

[b]"I'm trying to maintain a shred of dignity in this world." - Me[/b]

So let's see how Rosetta deals with the 5.3.2 when it's up again!

been running Rosetta on 5.3.2 for a few days now, and have had absolutely no problems whatsoever. not even the explorer.exe CPU issue that always plagued previous releases. I'm quite happy with it :)

---

been running Rosetta on 5.3.2 for a few days now

I'm confused. According to the BOINC website, the latest development version is 5.2.14 (use at own risk). Where does 5.3.2 come in?

---

*** Join BOINC@Australia today ***

been running Rosetta on 5.3.2 for a few days now

I'm confused. According to the BOINC website, the latest development version is 5.2.14 (use at own risk). Where does 5.3.2 come in?

Get it here

---

5.3.2

I'm confused ... the latest development version is 5.2.14 ...

It is indeed confusing... I'll attempt to explain the numbering system...

Odd middle digits are development-only versions. They are never supposed to be freely handed out or recommended, there's no guarantees they won't give negative credits, erase your hard drives, fry your CPU, and make you sterile. Only when they have been proven "good", after extensive testing, do they become "_even_ middle digit" releases. So...

Once V5.1.x had passed testing, it became 5.2.0 and was 'recommended'. Any bug fixes that are applied to 5.2.x become 5.2.(x+1). So 5.2.14 is the bug-fixed version of 5.2.13. It was posted as 'at your own risk', and some people were told to get it, before the testing was over, because it fixed the "connected to localhost" issue. Unfortunately, it failed alpha-test for other, unrelated reasons. It's a dead-end, will never become "recommended". There will probably be a 5.2.15 out within a couple of days that does what 5.2.14 was supposed to do, but without breaking anything else. Testing for a new 5.2.x version is limited pretty much to "did we fix the bug" and "did anything else obvious or major break", as changes from the prior version are fairly small.

Meanwhile, back at the ranch... Once 5.2.0 was out the door, the developers started adding in the "wish list" of new features and changes. That version was 5.3.0, and there's quite a bit of difference between it and 5.2.0. As each bug fix has been made to the 5.2 series, that bug fix was _also_ (hopefully) applied to the 5.3 series, wherever it stood at that moment. There is no "time relationship" between any 5.2.x and any 5.3.x version - 5.3.1 was compiled before 5.2.8. 5.3.2 was compiled after 5.2.13. On any given day, you have to have a scorecard to know which is the "newest" version - but it shouldn't matter, because nobody but the devs and "approved testers" are really supposed to be aware at all of the 5.3 versions, and they aren't using the 5.2 versions at all. Eventually the bug fixes will stop on 5.2, a 5.3 will be "blessed", and it will then be released to everyone as 5.4.0.

All of that said - BOINC is an open-source project. Anyone can, per the rules of the license, download the source to _any_ of the versions, and compile it themselves, so there isn't really much point in "securing" the "development" binaries either. The download site is, um, slightly more difficult to find, but that's about it. It is "legal" for someone to use any version they want to - it just isn't a good idea, unless they're paying close attention and know what to expect.

I generally don't mess with the 5.3 stuff, because I've been more concerned with having what everybody else is likely to have (for support) instead of worrying about "what's coming" right now. I hear good things about 5.3.2, but then it hasn't been tested very much yet. MUCH less than 5.2.13 has been... we at least know (some) of the bugs in 5.2.13. :-)

Now, the disclaimers; I'm not a BOINC developer, and not a BOINC tester. I read _some_ of the dev mailing list stuff, and the test web page forums, and look at the bugs database on occasion, but I am not "in the loop" other than as much as I make time to be, just out of my own interest. So while I think this is a fairly accurate overview, I'm not about to swear to all the details being perfect...

---

Yoda,

For what it is worth, I am now "standardizing" on version 5.2.13 as it seems to have fixed most of the other "teething" troubles of the 5.2.x series. I *HAD* been running version 4.72 up to this point as I had found it also to be decently stable (though it had a file-handle bug that caused problems with SETI@Home's troubles - to the point it was tripping off my wireless LAN).

Echoing Bill, I strongly suggest sticking with the recommended versions *ONLY*. As a matter of fact, I suggest some caution about just upgrading for the sake of upgrading (usually).

My personal recommendation is 5.2.13 and then stick with it until there is a compelling reason to upgrade. Then, upgrade cautiously. To make caution easier, we have an upgrade guide in the Wiki to make it easier ... :)

Following my own advice, I am just upgrading to 5.2.13 on my machines and should be doing the last two today. It takes me an hour or so, mostly because I also use the time to save off my old and completed CPDN models to my server (over wireless LAN that takes some time to save off 2-3 models).

Once upgraded, I will wait and watch. When I see a reason to try one of the new versions, I will try it on ONE computer ... if it causes problems, I back level and go back to a version I *KNOW* works.

I hope this helps ...

---

Yoda,

My personal recommendation is 5.2.13 and then stick with it until there is a compelling reason to upgrade. Then, upgrade cautiously. To make caution easier, we have an upgrade guide in the Wiki to make it easier ... :)

Following my own advice, I am just upgrading to 5.2.13 on my machines and should be doing the last two today. It takes me an hour or so, mostly because I also use the time to save off my old and completed CPDN models to my server (over wireless LAN that takes some time to save off 2-3 models).

Once upgraded, I will wait and watch. When I see a reason to try one of the new versions, I will try it on ONE computer ... if it causes problems, I back level and go back to a version I *KNOW* works.

That is exactly, why my 15 of 16 boxes are with 5.2.7; only my laptop is 5.2.14 (5.2.13 was not so good on the laptop)

---

Supporting BOINC, a great concept !

That is exactly, why my 15 of 16 boxes are with 5.2.7; only my laptop is 5.2.14 (5.2.13 was not so good on the laptop)

Exactly right!

What works for me may not work for you. I don't have laptops, my machines are close to the top in clock speed (though the 2.8 GHz machine is getting long in tooth and is slated for renovation when I am flush with cash again).

This is where having a couple machines comes in handy, you can try the upgrade cautiously then only start to convert machines if the test succeeds.

---

Echoing Bill, I strongly suggest sticking with the recommended versions *ONLY*. As a matter of fact, I suggest some caution about just upgrading for the sake of upgrading (usually).

Taking this a step further, if I'm running 5.2.11 across the board, and not having any trouble with it, there really is no reason for me to upgrade to any 5.2.x version where x > 11.

If it ain't broke, don't fix it.

---

Taking this a step further, if I'm running 5.2.11 across the board, and not having any trouble with it, there really is no reason for me to upgrade to any 5.2.x version where x > 11.

In fact, there is a valid reason for you _not_ to go to 5.2.13, at any rate. (There may still be some fix/change in the future, say 5.2.27 or something, that _would_ be worth moving to...) My Windows PC doesn't like 5.2.13; I get the "localhost connection" problem with it, that I did _not_ get with 5.2.anythingelse. I know how to get around it, it's not a big deal to me, and I "stay current" in order to be able to help others - but for anyone else, and within reason, yeah, "if it ain't broke..."

Now, that being said. Anyone running anything before V5.2.6 who is running SETI should seriously consider upgrading soon. And if Rosetta goes to flops-counting, V5.2.6 or above will be REQUIRED for this project. Only V5.2.6 or above have the necessary code in place to handle this.

There are still some running 4.19, only because their benchmarks were higher. Sigh. No scheduler, no retries on failed communications...

---

There are still some running 4.19, only because their benchmarks were higher. Sigh. No scheduler, no retries on failed communications...

No problem to stop this; set the minimum client of the project rosetta to a number that satisfies (uups spelling ?) the need of the project and they no longer get work; they only get the message, minimum-requirement to get work from rosetta is client-vers xx.xx.x

---

Supporting BOINC, a great concept !

...

Echoing Bill, I strongly suggest sticking with the recommended versions *ONLY*. As a matter of fact, I suggest some caution about just upgrading for the sake of upgrading (usually).

...

I agree with you, Paul, and I don't upgrade to any version just for the upgrade. If it works fine, there's no need to!

But I was allured by Ageless, who showed the new stats


Click the Pic!

and I asked him if it's stable, and he said it is. And I haven't had any problems with it at all. But again, I can only say it calls for caution to upgrade to the test versions, before they have been thoroughly tested out. Upgrading to versions above the recommended version is to take a walk on the wild side! And one should be prepared for unwanted consequences for one's system, and WU's can be flushed out in the process.

---

[b]"I'm trying to maintain a shred of dignity in this world." - Me[/b]