bitcoind runs without memory leak, but memory usage is high (1.2 GB on my setup, uptime 100 days now).

Ok, then it's fine... Just had a 60W running at 40W power brick fail on me after a year without being choked together like that... You're not even pulling 50% of max power on them then?

@Xian01 my guess is those powerbricks wont last 2 months like that...

Can you explain? Is there a way to add zeromq to bitcoind? Can you add a step-by-step tutorial?

We definitely need to add a long-poll option to listsinceblock, so it can block if result is empty.

ok, but 16GB is only going to last you a couple of months at this rate, the block chain grows roughly 1 GB per month. I got a 56GB ssd for the blockchain, if they haven't added proper pruning to the satoshi client by when that runs out; I'm selling all my BTC and ejecting.

btw, bitcoind 0.8.2 uses 1.1 GB RAM on my atom server, so still memory hog, that's not a problem when you have 4GB though...

About fullquote: talk to the forum programmer, he should disable it, it should take maybe 5 minutes. (if(post.body.length() > 500 chars || post.images.length > 0)) do not quote!!)

About heatsinks: My USB stick has now stabilized at 0.7% HW/A, which is 0.2% lower than without heatsink, both cases without fan! It would be very nice if anyone else could look at the PCB under the BE 100 chip and see if it is discolored!

Can you take a look under your bottom heatsink?

Can you take a picture of one of yours instead? There is going to be alot of hard words on the forum if these start to fail because of bad heatsink design. The problem is my chip had 0,9% HW/A witout HS, I'm pretty sure that would be higher if I add that new flat HS.

Edit: maybe it was premature, the HW/A is now up to 0.5%...

Still can someone look at their PCB?

I'm temporarily putting a fan on mine just to try and prove my case.

Ok, nm, still climbing, with fan, and two heatsinks... there seems to be no correlation between heat and HW?!

HW/A is 0.3% with old heatsink. I think we have a real problem! You guys should REALLY care, your investment is burning up!

There was photos of the old heatsink, but friedcat removed them. Maybe someone else has posted them?

Can anyone confirm that the PCB gets discolored by the heat with new heatsink and active cooling too?

If you make another GB at 0.99 BTC make sure you ask for the old heatsinks!

HW/A 0.2% so far!

So friedcat finally came through with the original heatsink. When installing it I noticed that the PCB was discolored under the BE 100 chip. This thing is definitely too hot, even with the new heatsink and active cooling I would say. Now with dual proper heatsinks, we'll see how it goes.

Il y a une tension qui se creait quand la difficulte et le prix est en dissonnance.

http://bitcoin.sipa.be

Je crois qu'on va voire une spectaculaire hausse bientot. Mais attention, ca ne veut pas dire que j'ai achete des bitcoin. J'invenstis uniquement dans l'equiment, c'est "plus" sure.

Not from the bottom though, what are you looking for?

I wish friedcat would send me one of the old heatsinks like he said... then I would!

It's not about size, it's about surface.

This little HS doesn't look like much, but it increases the surface by x9!

ASIC 6x6=36mm²

HS 8x8=64mm²x5(one for each blade)=320mm²

320/36=8,8 plus borders on blades and base = ~ x9

Why friedcat made the new heatsink without surface in mind is completely beyond me, specially since the original was!

Hm, no, but that's an idea; don't know how I would attach it though?!

HW/A is 0.8 again... we'll see after a couple of days. You can feel the air movement above the HS, I hope it's doing something atleast.

@pankkake Unscrew, it falls off.
@ct1aic It's from http://www.elfaelectronics.com, the heatsink is 8x8mm.

Will let it run for a while, I suspect same HW/A but longer life...