archive snowblossom slack mining
prev 2018-06-29 next
2018-06-29 01:15:01
Shoots
Bah @AlexCrow dropped off and so did our luck
2018-06-29 02:25:36
Clueless
@Fireduck now seeding snowblossom 7
2018-06-29 02:25:46
Clueless
if you want to chunk all of them, I can seed all of them
2018-06-29 03:10:41
Fireduck
It is a work in progress
2018-06-29 05:30:33
kaka
Are you kidding me?
2018-06-29 05:33:00
fydel
who is dwmve86sywjhk3xsfznj2wkjjv9j7rnc6z7afgwv?
2018-06-29 05:34:01
fydel
he started yesterday.
2018-06-29 05:34:33
kaka
@fydel Which website
2018-06-29 05:36:33
fydel
i put some munin scripts together.
2018-06-29 05:41:25
fydel
very simple (20 minutes of work) but very fancy. :rabbit:
2018-06-29 06:31:39
kaka
Too unfair
2018-06-29 15:01:29
AlexCrow
what is your HR on that setup ? so I used those flags and it didnt improve my hr on 1.1.2, butttt it reduced the memory usage on 1.0.6 so i can now run my 48 core vm that has 185gb of memory, should get better hash per $ now
2018-06-29 15:19:39
Shoots
3.3mh
2018-06-29 15:56:52
AlexCrow
I can get only ~ 2.7mh
2018-06-29 16:06:33
fydel
with the same machine specs?
2018-06-29 16:06:45
fydel
i am getting 3.25mh/s
2018-06-29 16:10:46
AlexCrow
I use gcp
2018-06-29 16:12:21
Shoots
with 1.0.6?
2018-06-29 16:13:42
AlexCrow
yes - 1.0.6 but 1.1.2 gets me same HR
2018-06-29 16:13:45
fydel
ah. okay. i am on aws.
2018-06-29 16:15:45
Shoots
some servers are better than others, anywhere from 3.25 - 3.35
2018-06-29 16:19:37
Rotonen
the variance has to do with how exactly the memory gets provisioned for you across the numa nodes and within node channels
2018-06-29 16:20:01
Fireduck
numa takin' about my moma?
2018-06-29 16:20:21
Rotonen
netflix actually has scripting which tests that at bootup and kills the instances if they did not get top bang per buck
2018-06-29 16:20:23
Rotonen
also for cpu steal
2018-06-29 18:17:01
fydel
i have set up hamsterpool for public. it has detailed stats and graphs.
2018-06-29 18:17:56
Fireduck
I love it
2018-06-29 18:18:49
Fireduck
you should lower than 1% to me, it is too much
2018-06-29 18:19:14
Fireduck
that goes for everyone
2018-06-29 18:23:04
fydel
okay. thanks. but devfee is good i want you to continue to work on this project. :wink:
2018-06-29 18:24:02
Shoots
1% is fine to the duck
2018-06-29 18:24:20
Shoots
Thats what we're giving as well. Great work hamster!
2018-06-29 18:25:26
fydel
hamster power! :hamster:
2018-06-29 18:37:38
fydel
okay.
2018-06-29 18:37:40
fydel
0.5% devfee for fireduck, 0.25% for fancy pool
2018-06-29 18:37:54
Rotonen
welcome to the sub 1%
2018-06-29 18:38:47
Fireduck
bam
2018-06-29 18:44:48
fydel
0.6% (0.5% devfee for fireduck, 0.1% for fancy pool)
2018-06-29 18:45:05
Fireduck
fee wars
2018-06-29 18:45:07
fydel
i just want to be a part of the party.
2018-06-29 18:46:07
Fireduck
I wonder what the cheapest board I can put a 970 pro in and get its full potential is
2018-06-29 18:46:58
Rotonen
@fydel...wat? :smile:
2018-06-29 18:47:08
Rotonen
it's not like i'm in the race or anything
2018-06-29 18:50:22
fydel
i just like this project and i want to contribute.
2018-06-29 18:51:41
fydel
i thought fireduck was encouraging to lower the fees to dev.
2018-06-29 18:51:55
Fireduck
@fydel I was
2018-06-29 18:56:01
Rotonen
i mean, there's no one you're actually competing against in regards to low fees - essentially no one mines on my pool
2018-06-29 18:56:14
Rotonen
but i suppose you're also the majority of your hashrate, so the fee does not even matter
2018-06-29 19:08:27
Shoots
in reality no one is mining on proto either, just the few of us associated with the pool
2018-06-29 19:09:44
Rotonen
the entry funnel from discovery to mining needs improvement
2018-06-29 19:10:29
Rotonen
i'm a bit headfirst deep into what makes the mining tick in general so i'm not yet in a position to start writing a polished welcoming narrative of any sorts
2018-06-29 20:13:45
stoner19
patiently waits for all of the AWS coupon ~abusers~ users to run out of money and nethash to drop
2018-06-29 20:14:17
Shoots
aws doesnt give you coupons or free credit
2018-06-29 20:14:29
stoner19
just a joke :slightly_smiling_face:
2018-06-29 20:14:31
Shoots
they let you run a couple micro servers that cant mine snow
2018-06-29 20:14:58
stoner19
damn, both of my miners stopped for no reason!
2018-06-29 20:14:59
stoner19
what the hell
2018-06-29 20:15:04
Shoots
once theres enough people mining and profit goes down it will level out
2018-06-29 20:15:09
Shoots
spot isntances?
2018-06-29 20:15:28
stoner19
no, VPS
2018-06-29 20:15:37
stoner19
I tried spot instances and amazon fucked me.
2018-06-29 20:25:25
Rotonen
you want to be running a cost aware spot fleet
2018-06-29 20:25:37
Rotonen
which suspends the servers when too expensive
2018-06-29 20:26:03
stoner19
but then you lose your snowfields and miner settings, etc when that happens
2018-06-29 20:26:03
Rotonen
as in this stuff: https://aws.amazon.com/blogs/aws/new-auto-scaling-for-ec2-spot-fleets/ The EC2 Spot Fleet model (see Amazon EC2 Spot Fleet API – Manage Thousands of Spot Instances with one Request for more information) allows you to create a fleet of EC2 instances with a single request. You simply specify the fleet’s target capacity, enter a bid price per hour, and choose the instance types that […]
2018-06-29 20:26:21
stoner19
so you have to have a snapshot or something from a "standard" aws instance to clone ?
2018-06-29 20:26:22
Rotonen
if that happens to you, you have no idea how to use aws, but YMMV
2018-06-29 20:26:40
Rotonen
no, you actually need to read their docs on data persistency
2018-06-29 20:26:42
Rotonen
and roll your own AMI
2018-06-29 20:26:43
stoner19
correct, I'm very inexperienced with AWS
2018-06-29 20:26:59
Rotonen
it's a relatively easy thing to learn and aws skills are in demand, happy reading
2018-06-29 20:27:09
stoner19
thank you for sharing.
2018-06-29 20:27:28
Rotonen
they have a *lot* of documentation, the hard bit is cutting through their marketing terms
2018-06-29 20:27:41
Fireduck
It is pretty easy, you basically create an ec2 instance, set it up the way you want it
2018-06-29 20:27:51
Fireduck
then you create an image of that
2018-06-29 20:27:57
Fireduck
which you can deploy as many times as you like
2018-06-29 20:28:09
Rotonen
@Fireduck as io bursts cost you extra, the pricing optimization is a bit nontrivial
2018-06-29 20:28:12
stoner19
deploy to spot instances because they're drastically cheaper
2018-06-29 20:28:13
fydel
but you need to find out where it is cheapest.
2018-06-29 20:28:27
Rotonen
@fydel not even that, you can just set a price ceiling for your bid
2018-06-29 20:28:43
Fireduck
right, you don't want to be screwed with while setting up the prototype image
2018-06-29 20:28:49
Fireduck
so don't make that one a spot
2018-06-29 20:28:58
stoner19
I already learned that the hard way :wink:
2018-06-29 20:29:03
fydel
yeah. but just for one region.
2018-06-29 20:29:12
stoner19
thus the reason I'm not using AWS right now :stuck_out_tongue:
2018-06-29 20:29:20
Rotonen
i'm pretty sure the bursts in global hashrate have something to do with the spot price of electricity at the utilities aws datacentres use
2018-06-29 20:29:21
Fireduck
GCP is easier about deploying images in other regions
2018-06-29 20:29:53
Fireduck
but GCP doesn't have the spot market, just just have a reduced cost for preemptiblew
2018-06-29 20:29:58
Rotonen
one thing with amazon is that their web clickety clickety stuff sucks and a lot of the advanced stuff is just broken
2018-06-29 20:30:10
Rotonen
so use a command line tool, or roll your own manager with something like bobo
2018-06-29 20:30:32
Fireduck
My experience differs there :wink:
2018-06-29 20:30:45
fydel
mine too. :slightly_smiling_face:
2018-06-29 20:30:46
stoner19
preemptible are 24 hours at a time max right?
2018-06-29 20:30:47
Fireduck
back in maybe 2010 you had to use your own tools since the web console was terrible
2018-06-29 20:30:59
Fireduck
these day, the console seems to have most everything working
2018-06-29 20:31:10
Fireduck
but maybe there is fancy stuff I don't even know about
2018-06-29 20:31:15
stoner19
meh, I'll get there eventually. maybe. would be nice to have some better hash on this, but I'm still grateful to at least have something.
2018-06-29 20:31:35
Fireduck
@stoner19 yeah, 24 hours but if you do an instance group it will restart them as there is capacity
2018-06-29 20:31:35
Rotonen
@Fireduck i've yet to figure out how to properly configure their managed kafka through the web
2018-06-29 20:31:58
Fireduck
@Rotonen getting ipv6 and ipv4 working was a bit of a trial
2018-06-29 20:32:10
Rotonen
but sure, as per past experience i'm just very easily dropping using their web thingy and falling back to the api
2018-06-29 20:32:40
Rotonen
@Fireduck i think i'm like 5+ iterations of their security groups bullshit behind, i just open the floodgates and do traditional iptables stuff
2018-06-29 20:33:10
Fireduck
security groups isn't terrible. It gets a little weird when it is your stuff talking to your stuff.
2018-06-29 20:33:19
Fireduck
Took me a while to get their new EFS working
2018-06-29 20:33:55
Shoots
I just got a script running to create aws instances, going to attempt to figure out how to scrape the explorer for the difficulty and start/stop them depending on the difficulty
2018-06-29 20:34:05
Rotonen
yeah, i'm a traditionalist so i want everything with publicly routable IPs, but still a segregated network - not what they hand in mind as the prime use case
2018-06-29 20:34:36
Fireduck
I like their setup, but yeah it is more complex than I ever want or need
2018-06-29 20:34:58
Fireduck
but if I were setting up a large/medium company infrastructure all on AWS I'd want it
2018-06-29 20:35:07
Fireduck
private network subnets, etc
2018-06-29 20:35:28
Fireduck
Google has similar stuff, but no IPv6
2018-06-29 20:35:31
Fireduck
which sucks
2018-06-29 20:35:48
Fireduck
Moved the snowblossom seed nodes to AWS to be on ipv6 as well
2018-06-29 20:37:59
Shoots
protopool is back baby @AlexCrow I assume thats you hammering it with hashes
2018-06-29 20:38:16
Fireduck
Is there anyone around here not named Alex?
2018-06-29 20:38:19
Rotonen
how is the protopool down that much? :smile:
2018-06-29 20:38:44
Shoots
I think we are somewhere around 700MH right now
2018-06-29 20:39:00
Shoots
based on my percentage of shares and my hashrate
2018-06-29 20:39:15
stoner19
I'm not Alex
2018-06-29 20:39:56
Shoots
we should automatically name all the anon wallets on the block explorer to "Alex" and then something generic
2018-06-29 20:40:01
stoner19
can we have protopool % decimal spaces increased to 8 so it at least looks like I'm a contributing member? :stuck_out_tongue:
2018-06-29 20:41:20
Shoots
@Protovist plans on adding more stats to the pool, he's just been busy
2018-06-29 20:41:53
Fireduck
ok, I should get back to work
2018-06-29 21:06:59
fydel
the problem with aws is to find out the instance/region which is the cheapest.
2018-06-29 21:15:08
Fireduck
Does it allow all regions or just all az within a region?
2018-06-29 21:17:34
stoner19
you can select a subnet/zone
2018-06-29 21:18:19
fydel
you are just in one region like canada, ohio or frankfurt.
2018-06-29 21:18:46
fydel
you are choosing the cheapest instance in your region.
2018-06-29 21:19:06
fydel
take a look at the text file i posted above.
2018-06-29 21:23:20
Fireduck
The new miner will be named Arktika in my continuing trend of naming shit weird things
2018-06-29 21:23:34
Fireduck
named after the russian line of nuclear powered ice breakers
2018-06-29 21:24:29
fydel
i am giving great advise and you are laughing at me.
2018-06-29 21:24:53
fydel
i think i will reduce dev fee to 0.01% :stuck_out_tongue:
2018-06-29 21:25:02
Fireduck
hahah
2018-06-29 21:25:18
fydel
:slightly_smiling_face:
2018-06-29 21:47:23
Fireduck
coming together in jeanluc branch