does cgminer restart every hour affect profit on Eligius apart from apparent 10-15 seconds no hashing time it takes for a restart? i.e. is it considered a pool hop, do I get less reward than if I would not restart?
I ask this because this is only way I seem to be able to keep my 12 hour AVG up on my lemon S3.

I bought from https://bitmaintech.com/product.htm
they really look different, one has curved heatsinks - the bad one (just like this http://img.youtube.com/vi/VOyrCut0y30/0.jpg), other has straight - the good one. If you are still interested, by the end of the week I will take pics for comparison, I have decided to run them until end of the week without restarts (so will not remove cover - this will require downtime) and see what AVG I get. After I took them down for testing yesterday they both had good 3 hour AVG during night (890 for both), and now it is 5% down as usual for one of the units and going lower (now it is 865), I have a feeling it will stay there and go lower - to 840. I will remind in pm when I have full analysis to those who were interested about this. EDIT: "3 hour AVG for two miners not 865, but 840 already - I think this is what it will stay until I reboot on weekend, so 5% drop for both units, but actually only one is losing speed, which makes it around 10% for that unit"

I only cleaned heatsink on the bad one, it was covered in thick dust layer at the front side (probably the good one is the same, but aint touching it while it runs good), it runs more quiet now after cleaning, but the hashrate AVG behavior did not change.
I already contacted bitmain, they suggested: firmware update (which I did not do, because from what I saw it brings only features I do not need and from what I read for some people it got worse after update), and testing hashboards one by one by disconnecting them which I did not try also.
I agree there is no point in sending back if it is 10% slower than it should be (or rather 5%, because bitmain states that -5% is ok). I was hoping that restarts are good idea, but it seems not - losing shares?

I also tightened screws on heatsinks of the bad unint as one of posters on this page recommended.

So lets wait until I post analysis of running without interruptions and thank you for the interest, did not expect this, my initial idea was just to ask if the restarts is a good idea

ok, thanks for reply, I will bite the bullet then and let it run at whatever hash rate it feels like without touching it

I only found out about Bitcoin a month ago, I only have 2 S3s for experimental purpose, I am not expecting them to ROI (ordered late (batch 6) and bought coins while they were 600 a pop). using them as heater for appartment, it is cold here already - 18 degrees during day outside.
yes, tried power supply swap. tried moving the units apart, using different power outlets, swapping pci-e ports (now they run all 4 pci-e ports connected) and what not.
the hash rate does sometimes (like the one 30 hour period I mentioned) drop 10% which is below the expected rate.

thank you for looking into it, I think this needs engineer to find the cause of the problem

here is someone reporting same behavior:
https://bitcointalk.org/index.php?topic=671189.3900
https://bitcointalk.org/index.php?topic=671189.msg8169759#msg8169759

edit: this is offtopic anyway, my question was - does restarting cgminer every 1 hour cause problems with pool shares?

I am 90% on Eligius, but have tried Slush, Ghash, Discus Fish - all of them report same behavior.
The only pool I am getting stable hash rate on bad unit is solopool.net. Only difference I can see is that most pools will set difficulty to 256 or 1024, but solopool is setting it to 50, but I do not feel lucky enough to mine there.

I am not taking antminer status page that much into account, but am looking at pool stats.
I have been running miners for 2 weeks now. The problem is with only one of the units, they both are not similar, they have different heatsink design, fans on good one are really loud and running 1980RPMs AVG one temp sensor shows 40, other 35, the bad one is very silent and fans run at 1620AVG, both temp sensors show 38 degrees.
when I first started the bad one, it ran at 440 GH/s avg for about 12 hours, then took a dip to 400 AVG for 8 hours and back to 440 for 3 hours. Then there was a power outage in my area for 1 hour, after power came back it started hashing 400 AVG for 30 hours non stop. I came home and tried to reboot for several times, it did not help (10 minute AVG was showing 400-420 GH/s). Then tried to reset to factory defaults, which fixed the issue for 6 hours hashing at expected rate, then it went back to 400 GH/s AVG (for no appearent reason). One morning I reset to factory defaults again, it hashed at 440 GH/s for a short period of time (30-60 minutes) and then it was back to 400 GH/s for 8 hours. There are no X showing in miner status, temperature is lower than the other miner I have (which runs 440 avg no matter what). Acceptance rate / rejects ratio are very similar. After seeing it run slow for 30 hours non stop I am not letting it do this anymore, instead I am trying different things with it - one of which is auto cgminer restart.

edit: by the way, I have found at least 2 people on this forum reporting this same exact symptoms as I have and are auto restarting as well.

my 2x antminers s3 batch6 seem to be running very unstable for me, when I first start them cold they run fine for many random amount of hours or days, and then suddenly losing 6-7% of power (from 440 Gh/s to 410 Gh/s), I have been experimenting alot, cleaning the heatsinks, running open/closed case, reversing both fans, keeping room they are in as cool as possible, but they seem to be doing that no matter what I try. restarting cgminer seems to be keeping AVG hash rate up.

lets say I have set miner to restart every 1 hour, does that mean the mined shares will be lost?

you should understand - if that happened too many people (holding just a few BTCs) would become rich and value of money would be lost.
on planet earth system works in this way - 90% of whole people are poor and supporting 10% wealthy.
it is impossible for 10% to be poor and 90% wealthy, this would bring value of money down and again would benefit only 10% that are the richest.

or you think majority of people on planet earth can be rich in a couple of years from now and live happily ever after, most people owning yachts, expensive cars big houses? Where would you build so many houses for every rich person and who would build the houses if most people were rich?

I saw the same thing, went back to check at location, miners are online and hashing.

blockchain.info is showing 3 blocks in last 6 hours:
317254 (Main Chain)    2014-08-24 10:19:49    00000000000000002bf5c1b98e4686f9675b6a7d4c00e94b5aa2ed3463339b8f
317248 (Main Chain)    2014-08-24 09:28:38    000000000000000009965f11eea44470f7122ba63535d7b74cdd9bd95d1239a2
317228 (Main Chain)    2014-08-24 04:31:07    000000000000000029f8ddca13878ddd9c9b616b91238bd28c474cb1deae57cb

but Eligius stats home page shows two:
12 minutes   2014-08-24 04:32:54   05:48:45   48,542,819,584   23.84G   49.1%   9,963.67 Th   2 of 120   12igyTG1TB...   317,254   ...4b5aa2ed3463339b8f
6 hours   2014-08-24 00:40:29   03:52:25   32,256,559,360   23.84G   73.9%   9,934.81 Th   28 of 120   19chYFuCv1...   317,228   ...d28c474cb1deae57cb

is this blockchain.info not showing correct info or is it Eligius stats page, regarding block:
000000000000000009965f11eea44470f7122ba63535d7b74cdd9bd95d1239a2
?

EDIT: I now understand that I was wrong to ask this, because actually blockchain.info can not possibly know for sure where the block originates from, it is just a good guess, more info here:
https://bitcointalk.org/index.php?topic=123726.0

make it 15blocks/24 hours

for 20 000 PXC you need 800 KH/s, which is almost half of whole network hashrate, I would be solo mining with that much power myself if I had that

I found out about bitcoin a month ago, it is now 1 week since I started mining, I was one of the miners hopping after seeing slush luck going up to the moon, I already knew my S3s are not going to ever ROI since I started mining (bought batch6 - late, got BTCs while they were high priced), this is just an experiment for me and love doing this and learning while at it. might even buy "next big thing" hardware and by then Ill know most DOs and DONTs.
It is hard for newbies like me to resist to try something like hopping.

same here, you can easily check if site is down using something like this:
http://www.downforeveryoneorjustme.com/http://eligius.st/

edit: it is already up

if you look at top of the current queue: http://eligius.st/~wizkid057/newstats/payoutqueue.php

You see that at very top there are people who have not been payed for 6-5 days, so if you at this very moment would have not been payed for 7 days (i.e. it would now be a week since you last got payed), then you would be payed as soon as next block was found and you would be in front of all those people who have not been payed for 6-5 days.

So in other words - if you set out to be payed every, lets say 0.08-0.1 BTC, you would have to wait for that amount to be earned for longer time, but would get payed without waiting in queue.
In average it does not change anything, you would still be payed every 6-7 days if you have configured it or not, but only difference is psychological, you would know, that when you have 0.08-0.1 BTC after 7 days, then you would be payed instantly.

I was also 6-7 blocks until payout ~3 days ago. I think you should also expect another 3-4 days until you get your payment

not too small, it does not affect payout queue. you should calculate how much you will be payed after ~7 days and set you personal config on Eligius to pay out that ammount as minimum - no queues then

#47 should be payed out after this round is done - 1 block is found. or are you in top ~50 for a long time?

did you reach a top of payout queue ( http://eligius.st/~wizkid057/newstats/payoutqueue.php ) and did not receive a payment?
I am supposed to get my first payout after next round. I also have set up to be payed out every ~7 days (or 0.1 BTC) to not sit in queue anymore.

I think it is impossible to be flat 0%, it can be very close to it if everyone stopped hashing at this pool for a very long time, lowest I can find for this pool is 10.7% back in 2013-11-07.

But there - we are done with previous round in 10h:17m:50s.
This is fun!

but we are at streak of bad luck indeed, I got 40% shares shelved, started mining 4 days ago.