Again same thing, can't place or cancel any orders! Last trade timestamp was 30 minutes ago!
Why does this keep happening?

Can log-in, but can't cancel trades, can't open trades. Last timestamp at cryptowatch is 20 minutes old!
Status page shows like there isn't an issue. The hell is going on?!

Website uptime 0%!

Is this some sort of reenactment of 2017? I've been sitting here the last hour pressing F5, and the best I got is a blank trade page FFS!
Clearly someone is still trading according to cryptowatch, so which fu**ing side of the planet do I need to fly to cancel my orders?

My order was a limit buy set higher than what market prices showed at kraken site and cryptowatch charts (all my order are limit orders).
Setting market orders @ kraken is like playing Russian roulette with a semi-automatic handgun.
Yet for 2 minutes 41 seconds the order remained untouched while the price kept ticking lower than what my limit buy was set at according to both sites!

After the usual delay of several minutes with all the errors for the site to finally open a buy order and list it in the order book, I watch the price refresh and tick lower and lower...
Finally 2 minutes and 41 seconds after listing the order in the order book, it executes, when the sells are over 90 euros lower than what my buy is set at! Pure comedy!
The order book refresh rate @ cryptowat.ch is next to useless when it matters the most. I lost count on the times it completely stopped refreshing, leaving me blind.
And with the excruciatingly slow refresh rates when it DOES work, even krakens own tab header shows more accurate numbers, that is when it refreshes once in a while.

After multiple errors while trying to set a limit sell order, which finally executes after the price already dropped lower than what my limit order was set at, I see the
price reverse back up and try to cancel my order multiple times in the spawn of 3-4 minutes that I had, and yet failed to do so thanks to error after another. Finally I get
"Cancel Pending" which grayed out and disabled the cancel button... 1 minute later it still shows cancel pending?! What the hell is cancel pending, it either cancels or
doesn't! And it didn't! I refreshed the page and see it sold the coins and I couldn't do a damn thing about it.

http://oi65.tinypic.com/nnmanl.jpg

Since today's dump, it now takes up to 45 seconds for the page to refresh from one cloudflare error into another!

Lights went out again at the cesspool @ 14378.70, right in the middle of the second dump. This time without cloudflare errors, just infinite times:
Your order may have failed. Please refresh the page and check your orders. Contact support if necessary.

Today Kraken gained a new feature - Invalid Request You must Login to perform this action.

How am I supposed to make anything out of the charts when it's filled with bs like this:


http://i67.tinypic.com/vevs7.png


P.S. This is getting more frequent

Unlike v2.1, this one hasn't crashed so far, but again 280x with elpida doesn't seem to be running right after a while, as I checked, I saw these anomalies http://bayimg.com/AAOieaafM and http://bayimg.com/aAOIgAafm

As others have already pointed out, v2.1 stops submitting shares after a random amount of time, and unlike previous version that stopped functioning only on my 280x's not 290's, this one stops working with all cards in a random amount of time (sometimes it won't even make it 10 minutes) Not only does it stop submitting shares, but it leaves the clocks at high powered state, i.e. it's drawing same amount as if it was mining, yet nothing happens. Sometimes -wd 1 kills the tasks, sometimes it doesn't. Cmd box shows last share submitted msg and nothing else after that. No error messages, gpu drivers haven't crashed, and the miner client is still responsive, but does not generate any msgs - no "new job received etc, nothing, nada, zilch (pools also show zero hash rate)
In this sate basically all it does is piss away electricity at full speed, like a locomotive with no driver and no destination.

I don't think cards crashing has anything to do with a powersupply, because:

A. The algo draws about 40% less watts, compared to scrypt, and I don't experience crashes with scrypt, xpm, keccak, etc.

B. I never caught the 290's crashing in the one week period that I mined XMR.

C. I spotted 280x with hynix crashing twice so far in that weeks period, while 280x with elpida ram might hash anything from a hour to six before it finally drops to idle clocks and does nothing but wastes energy.

D. Clocks have no effect to the rate of crashing (unless absurbly high, duh) I can set it bone stock, oc, or underclock and it crashes all the same.

My conclusion in the brief time I mined this coin - elpida 280x cards crash all the time, and have worse hashrates than their hynix counterparts (and thats even clock to clock)
Profit estimates and what you get in reality, are far off. Even when the pools aren't constantly down from ddosing. I don't know why, but I'm guessing a whole lot of shares get blown in the wind, thanks to wonky pools that don't seem to find shares, or they get orphaned, or what ever else is happening under the curtains.

Overall mining XMR with crashing elpida cards seems to be a chore of babysitting the client, or risk comming back to check on it the next day, only to find out it's been sitting idle whole night!
Oh and the current gui wallet seems to be more like a duct tape do-it-yourself, for yourself, rather than something for public use. (I'm used to Qt wallets)




PS: The wallet synching, while wastly improved, compared to the ridiculous time it took to get 24h's worth of updating, still makes my internet connection totally unusable.
Redemption to this was to download a bandwidth limiter and choke the upload link, that it totally saturates. Of course limiting the download speed isn't such a bad idea either.

I also more than not, see a lower hashrate than what the actual miner shows ( actual hashrate according to miner: 1605~ give or take 10/s).
I would guess this is just because the diff is set so high, that the shares come in at such long intervals, making the estimates drop. Still it should show a hashrate considerably over your actual speed every once in a while. Which it does for me, but rarely (I see 1.15 - 1.30KH/sec majority of the time, it even goes as low as 0.75KH/sec). Also the pools I tried, like moneropool.com among others, are frequently disconnecting and reconnecting all gpu's in random intervals ( every 5-40 min). Albeit not all at once, so I would think the issue is not my connection but a problem on the poolside/miner client.

To me something seems off with either the profit calculator estimates, pools, the miner, or everything combined. Since I'm hashing at 1600h/s, yet every day I receive 30% less than what is supposed to be. Even with the miner fee and pool fee + lost shares, that can't be 30-35% less? Also it seems stock elpida 280x cards suffer from poor hashrate (260~ vs 364 hynix) that is a considerable loss, just like in scrypt coins, and overcloking the elpida cards doesn't seem to bring more than 30h/s more, still far off from a otherwise identical 280x with hynix vram. You can OC the gpu's considerably higher than say scrypt coins, yet overclocking seems to have very little effect to hash rate and power consumption on a 280x & 290. Seems like some parts of the cards are hardly utilized at all.