TPU's WCG/BOINC Team

[80-watt Hamster]

der ain't no got dang jerbs

Anyone else keep running out of WUs? The 1920x seems to get the most for me but I keep seeing everything idle.

Yeah. I'll have a full queue for a bit, then go mostly or completely dry for an indeterminate amount of time. Also frequent comm blackouts. Have two crunchers running ATM; 11700K box seems to get far more assignments than its 5600G compatriot.

 

[windwhirl]

Received a bunch of MCMs yesterday night/this morning. I also had a few older WUs aborted by the project

 

[80-watt Hamster]

I've been assigned tasks, it'd be nice if they'd download, though.

 

[windwhirl]

I've been assigned tasks, it'd be nice if they'd download, though.

View attachment 261087

Ah, so it wasn't just me. I have these ones sitting around, waiting to finish downloading since... who knows when

 

[Toothless]

This is so dumb. I cancelled off the Rainforest project so the smaller WUs would come in easier. Still have a two day wait for ONE small unit.

Big reee.

 

[Zachary85]

Just Lovely. Looks like world community grids https certificate has expired and the site cannot be accessed. Can't even connect to send back completed work units now.

Update: A post on Twitter now says they are aware of the certificate issue and are working to fix the issue.

 

[debs3759]

Glad I have a 24 hour queue, makes the certificate issue less of a problem - assuming they fix it quickly.

Just set my client to no new downloads, as swapping my boot drive to an NVME, and that'll make it easier when I have no work left.

 

[Toothless]

I had less issues before they moved. I'm not understanding why the project backoff is happening so much but this is getting old. I found if I play the game if clicking "Retry Now" I maybe get a few jobs. Except the Titan eats through GPU tasks in a few minutes so that's just, a thing.

 

[phill]

Apologies guys I've been not much around but been having a load of problems myself trying to get work down and now uploaded... Thank you @Zachary85 for mentioning the issue/fix from Twitter... Honest to god, if I'd had known having a faster internet was the cause of all these problems.........

Had the 3970X on the go today, loads of damn work units to upload but nothing going anywhere and the same goes for anything to download... Guess they really do know how to fudge something up all at once... Wonder if this will carry on after they get all this sorted or will we be in for a time without problems??........ Anyone putting down betting odds??....

 

[stinger608]

Yep, as @Zachary85 stated, I'm not able to upload units as well. Jeez, every since the move it's been one thing after another.

Wish they would get everything hashed out so all of us can get back to some serious crunching.

 

[phill]

Yep, as @Zachary85 stated, I'm not able to upload units as well. Jeez, every since the move it's been one thing after another.

Wish they would get everything hashed out so all of us can get back to some serious crunching.

It would be great mate that for damn sure.... Over here we need to sort out the electric/gas prices but there we are!!

Hopefully with this fix, their points might actually start working again also....

 

[windwhirl]

Just to add further confirmation, I'm also stuck with a few units waiting for upload or download.

 

[stinger608]

Yep, same here @windwhirl

Been that way since yesterday some time. Sucks.

And, when I update it just gives me "communications differed 3:28:32. And starts the count down. If I update again, it just gives me the same dang thing.

 

[debs3759]

Realised that EaseUS Todo Backup can clone disks, so have moved my boot drive to the NVME disk. Not sure how much difference it'll make in speed, but it's double my old boot disk for similar cost to a SATA. disk. The SATA SSD will do for a scratch disk and swapfile (not that I need one), etc. Probably useful for temp downloads as well

 

[phill]

I think guys they might have solved the issue with the web and possibly the uploads.... Whilst checking for the stats today, the WCG site was opening..... Could it be too good to be true??......

 

[mstenholm]

I think guys they might have solved the issue with the web and possibly the uploads.... Whilst checking for the stats today, the WCG site was opening..... Could it be too good to be true??......

The Free-DC numbers are from yesterday up to the time when their certificate halted all communication. This is based on what I managed to produce and upload. All is back to normal, i.e. slow download with the added slow upload. I cleared my queue and while doing that got and completed a few OPNGs this morning.

 

[Zachary85]

The certificate that expired yesterday still showed Infernal Business Machines (IBM) as the certificate holder. The new certificate has had the name changed and no longer shows IBM on it.

 

[phill]

The Free-DC numbers are from yesterday up to the time when their certificate halted all communication. This is based on what I managed to produce and upload. All is back to normal, i.e. slow download with the added slow upload. I cleared my queue and while doing that got and completed a few OPNGs this morning.

Its the usual poop show I'm sad to say... I've got 6 machines whirring away since the weather is good here for 5 minutes and nothing but not connecting and downloading in... So frustrating... What have they done or are they doing about the problem I'd love to know....

 

[debs3759]

I woke up this afternoon to find the certificate issue fixed, and probably 36 hours work queued (I have it set for 24)

Will be setting up a 6C12T (ES CPU, similar to i7 8700) CPU this coming week. Not sure when I'll be able to set that up in the device manager (should be able to handle 3 ARP, just hope I have some spare DDR4 memory for it, or the i7 6700K will have to be temporarily taken offline).

Still having to manually retry downloads every few hours though, so it's not perfect yet

 

[phill]

I find I'm forever playing the game of how many button presses for retry download it takes to get some of the work units downloaded... Even with my massively slower connection I never had any issues with any of the downloads, so I'm wondering if IBM and Krembil never really spoke about what was going on... It seems that the upload is fine, but the download is just shocking... Here's hoping for some better results soon

 

[debs3759]

Just bought 16GB 3200MHz DDR4 for the Coffee Lake system. Going to try Win 11 on it, as it's my first system that should support it. It'll be my new daily workhorse as well. Not sure whether the motherboard is OK, as I bent a few pins then straightened them as best I could. With 8 threads on the 6700K and 12 on this, I should be able to generate 16K points a day. Hope to find a way to fund a Ryzen9 5950X in a few months as well. Got a few Noctua NH-D14 and an NH-D15 to keep things cool.

 

[Toothless]

Just bought 16GB 3200MHz DDR4 for the Coffee Lake system. Going to try Win 11 on it, as it's my first system that should support it. It'll be my new daily workhorse as well. Not sure whether the motherboard is OK, as I bent a few pins then straightened them as best I could. With 8 threads on the 6700K and 12 on this, I should be able to generate 16K points a day. Hope to find a way to fund a Ryzen9 5950X in a few months as well. Got a few Noctua NH-D14 and an NH-D15 to keep things cool.

5900x is hitting around the $400 mark, and I can tell you now a 5950x needs some serious cooling if you want PBO on.

 

[NastyHabits]

Late night update from California. After watching one unit run while downloads wait forever, I checked my PC's before I headed to bed. One of them is back working as normal - WU's running on all threads and twice as many waiting to start. Hope springs eternal, and is sometimes rewarded.

 

[Toothless]

Looks like downloads for me are working again. Probably just jinxed myself but I'm just happy I have WUs again.

 

[stinger608]

Yeppers, they must have gotten the issue sorted out.

Have a ton of work units ready to run now.