Ok I think I have far too many tasks on my main machine 6008784.

I've come home to find the GPUs revving down & BOINC Manager 6.12.34 is unusable - all the tabs are empty apart from Projects & click any button & it just sticks at "communicating" message. The cache has been large but now seems 'too' large. I have reduced the prefs to 5 days from 10. Should I let it just work its way through it? GPU driver 296.xx monitor not powering down.

Client_state.xml is 14MB. Files are updating in explorer - I just don't have the GUI element at this time. Is there clearily a limit to how many tasks BOINC can handle?

GPUs seem to be up & down. It is of course warmer today but I can set the GPUs to fixed fan speed if required. Anyway temps are within normal limits anyway.

(What are http_temp files?)

Do you have the manager set to only show only active tasks or all tasks?

For larger caches it is usually recommended to only show the active tasks.

---

Ok I think I have far too many tasks on my main machine 6008784.

I've come home to find the GPUs revving down & BOINC Manager 6.12.34 is unusable - all the tabs are empty apart from Projects & click any button & it just sticks at "communicating" message. The cache has been large but now seems 'too' large. I have reduced the prefs to 5 days from 10. Should I let it just work its way through it? GPU driver 296.xx monitor not powering down.

Client_state.xml is 14MB. Files are updating in explorer - I just don't have the GUI element at this time. Is there clearily a limit to how many tasks BOINC can handle?

GPUs seem to be up & down. It is of course warmer today but I can set the GPUs to fixed fan speed if required. Anyway temps are within normal limits anyway.

(What are http_temp files?)

I find more than 1000 tasks on a machine slows the GUI down depending on the machine. So I don't run the manager on most of my systems. Which makes it a non issue.

In my starts up for BOINC I have it delete all of the http_temp files. IIRC they are temp files from partial up/downloads.

---

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

I believe it was showing all but I have it clicked away from that tab anyway as I'm aware it can slow things down. But when tasks were visible when I first came to the machine this eve, clicking did nothing to update the manager visually. I haven't. Just wondering if it just needs to burn work off a bit until the cache lowers. Would like to click update but can't click anything (yet) ;). Is it worth reinstalling BOINC?

I believe it was showing all but I have it clicked away from that tab anyway as I'm aware it can slow things down. But when tasks were visible when I first came to the machine this eve, clicking did nothing to update the manager visually. I haven't. Just wondering if it just needs to burn work off a bit until the cache lowers. Would like to click update but can't click anything (yet) ;). Is it worth reinstalling BOINC?

I've had something like this with 6.10 ad also with 6.12... even with much less tasks... Ive found that if I exit the Boinc Manager without shutting down the client, and then restart it, it gets "conected" and every tab worked...

---

Ok I think I have far too many tasks on my main machine 6008784.

I've come home to find the GPUs revving down & BOINC Manager 6.12.34 is unusable - all the tabs are empty apart from Projects & click any button & it just sticks at "communicating" message. The cache has been large but now seems 'too' large. I have reduced the prefs to 5 days from 10. Should I let it just work its way through it? GPU driver 296.xx monitor not powering down.

Client_state.xml is 14MB. Files are updating in explorer - I just don't have the GUI element at this time. Is there clearily a limit to how many tasks BOINC can handle?

GPUs seem to be up & down. It is of course warmer today but I can set the GPUs to fixed fan speed if required. Anyway temps are within normal limits anyway.

(What are http_temp files?)

Wow! I had to comment when I checked your tasks in progress. 8,152! That, Sir, is one S**tload of a cache. I'd be inclined to reduce my cache, as well, since work seems to be flowing pretty steadily of late. Even my twin 590 rig is only carrying around 5000 in cache, with a five day setting. And yes, with a cache that large, the GUI will severely choke. Found that out from personal experience.

---

Well. Release the Krakken, eh?


The servers shall give as ye ask. Finally.

Now those who wish to get what they really want must fix their settings.......
LOL.

There may be some side effects lingering from the change in WU attributes.

Or the version of Boinc you are running.

Or whatever DA is up to this week...LOL.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

Reduced to 5 days - may reduce more. To say 2. CPU work is steady - always has been. Is the CPU so busy updating things it can't feed the GPUs?

Also my job_log is 20MB after a year...!!!

Since work units have been going pretty well I dropped from 5/5 to 5/2 on my cache settings. I've got ~3700 tasks on my main machine. GUI seems fine, it was as high as 6000 and never even missed a beat or ran weird.

---

Traveling through space at ~67,000mph!

And Manager's come to live again now. Cache 2 days.

And Manager's come to live again now. Cache 2 days.

The server is still showing you have nearly 8000 tasks on the machine. In progress (7998) Perhaps you just needs to keep it under 8000?

---

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

Something like that. It'll work itself down now with the smaller cache. Think we've accidentally reached a limit here :). Talk about pioneering... Pioneering to the point of strangling itself. GPU fans/temps more constant.

Something like that. It'll work itself down now with the smaller cache. Think we've accidentally reached a limit here :). Talk about pioneering... Pioneering to the point of strangling itself. GPU fans/temps more constant.

I recall some of the mega GPU crunchers limit the number of tasks they get with the disk limit settings. Deepening on if BOINC counts bytes or kbytes something in the rage of 2.8GB-3GB would probably work for that machine. I use a limit of 1000GB in BOINC even though across my 30 some odd machines I only use about 3.5GB for all S@H projects.

---

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

Good idea, it's over 3GB yes.

I use a limit of 1000GB in BOINC

1TB? You are setting yourself up for a big fail if BOINC goes bonkers and starts downloading a ton of work.

---

Think that's a typo ;)...

No I have BOINC set to 1000GB. 1TB would actually be 1024GB. The vast majority of the machines I am running don't have anywhere near that kind of space on the partition where BOINC runs. BOINC will run out of room long before the software limit applies. I was using 10GB, but I got tired of seeing things like "Free space 45GB" "Free Space available to BOINC: 9.95GB".

My 24 core box with ~2000 tasks:
Total disk space 2 GB
Free Disk Space 1.19 GB

---

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

Going down... <ding>

Update: not out the woods yet - GPUs just "barfed" (I got that word from you lot) just when I turned monitor on & now it's a case of don't-open-manager.

Going down... <ding>

What i do to try and stop getting too many tasks (and to keep my cache long) is every day or so, set NNT, then force Boinc to jump into High Priority by putting 14 days into the 'connect about every' local preference,
I'll start a good collection of CPU tasks, then suspend them, then remove the 14 days local preference, I'll let Boinc complete that selection of CPU tasks (while trying to make sure no more than two shorties run at once),
then rinse and repeat, Once all the shortie CPU tasks are done, I'll leave the GPU doing it's shorties, my GTX460 can do ~30 shorties an hour, your mileage may vary especially if you're running multiple tasks at once.

Claggy