Topic: [ANN][BURST] Burst | Efficient HDD Mining | New 1.2.3 Fork block 92000 - page 1206. (Read 2170889 times)

This was not a mining block reward but looks like a transaction was sent on that block.
EDIT: This seems to be my own transaction of 10029 burst (was playing with the wallet).

So, what just happened ?

EDIT: ah, just found out this was just sent along the block!

#2637 now, wonder who has got the block #2636.

Congrats to whomever cracked that one. WOW 

Its almost an hour now right?

Yeah, I was wondering if my client was stuck at first. Just another 65 blocks until the new algo kicks in.

Yup! It's taking like 4ever!.

Wow. 2636 is taking a LONG time

NP. Just let us know. Its very very likely that future users will encounter similar issues so we should get it figured out.

Haha, agreed. After doing this I thought that a better idea would have been to put it on a VM... but it's already done now, so whatever. I think it'll be fine once the initial plotting is done, then the resources of the miner are very low.

I will let you know what the next crash on the pause results. Thanks!

I don't know then. All crashes I've seen have been directly related to ram starvation.

If there is nothing obvious then you will probably have to wait for the pause to catch the error.

with what I can gather it sounds like you are pushing it on a pretty important server. (DNS etc) just make sure your not pushing it to hard 

Memory amount on that server is 32GB, currently being used is 57% of that total, the DNS server being the biggest utilization.

However, this server did crash last night and forced me to reboot it. Therefore clearing the memory. The miner still crashed one time after that restart.

My best guess is that it has something to do with the virtual machines running on the system, but I am not sure.

ALSO....

I checked the event viewer and there was nothing java related at all. I didn't want to keep quoting the other quote, as it was getting far too large and soon it will take a page per reply. lol.

How much ram do you have? If you're running too many jvms with not enough ram on windows, sometimes windows pops a box asking you to close something, then if you don't it auto-kills something. I have a comp with 4G ram running 2 miners that tends to kill 1 miner on occasion, and on another comp my wallet seems to have gotten auto-killed about when I left for work today, causing me to lose about 9-10 hours of mining time.

Best suggestion if lack of ram is the case is lower the -Xmx values, or add ram.

This can also be triggered by scheduled services starting up, which increase ram usage, so check the scheduler if anything is set to run.

never heard of windows storage space but may it be a simple timeout due to systemload?
never tried running the mining on windows due to the 2tb volume size limit but it sounds like microsoft has broken this with their new storage space technology.
if the drives are attached by usb3 may it be a cable issue? i had struggled some time ago with a 28 port usb3 hub. some attached devices have frozen the whole bus and forced a port reset by the driver resulting in loosing all devices.
it was one of the cables which broke everything...
if you have the plots on the failing drive already generated try to plug it into a usb2 port. the performance should be enough to parse 8tb plotdata within 4 minutes.

Alright, I have hammered down pricing and the backend infrastructure for a hosted Burstcoin mining solution.

Here's the pricing:
0.06BTC gets you one month mining with one terabyte (after formatting you get a full terabyte). Each subsequent month is 0.03BTC. The initial high cost is to account for generation, which requires significant computational resources. I have no idea about profitability, and it is far too early in this coin's life to predict anything. For now, of course, generating and mining with a plot will require your private key to be held on the remote server, so the smartest method would be that I give you the public address that is being mined to, and twice daily I empty the mining address to an address of your choice (in this way you can see the mining occurring publicly, and twice daily all mined balances are secured in an address only you have access to).

This service would also work well with pooled mining upon release. Discounts are available for larger orders, with 10TB+ being priced at 0.055BTC/TB, and 20TB+ being priced at 0.05BTC/TB. I have 3TB ready to go live (as a bundle, as the plots are made for one address), and can put extra Terabytes up in a timely manner.

Each server is based on 4TB hard drives, which when formatted result in 3.6TB of space I can sell.

Start with the application logs. It will probably be Java related.

Event logs are your best friend when you need to diagnose a windows issue. Its a good idea to just dive on and see if you can start to understand what problems are occurring. Then you can google the error codes and event ID's and get answers for the problem.

This strategy reminds me of dynamic disks in windows server. there was always some random issue with those that made it a bad idea to use that way. But maybe the found a way to stabilize it in windows 8. Haven't tried it yet. Looking forward to hearing how it turns out.

Not at this point. But what exactly would I be looking for. There is so much there that it would take a very long time to look through without knowing a general idea of what I am looking for.

This is a really decent idea. I may have to try this later.

yes, thats true (good point!), if you combine multiple drives to one big drive you have no problem ... but can windows handle it, if one of this drives fails ... will i lose all my plots?! anyway ... too late for me right now :-) but all others with same problem now have 2 options :-)