Topic: Phoenix 2 Miner v2.0.0 - page 2. (Read 201568 times)

hm, with cgminer I get even less Mhash than Guiminer.

I don't believe phoenix is supported anymore.  I suggest switching to cgminer.

M

After around 1 minute this happen and gpu stop minning.


.cfg

Same result running CMD as admin... Ugh - frustrating.

Try running cmd as an Administrator. Connection to the pool doesn't seem to be a problem. You would get something like that if there is no connection to the pool:
[10/26/2012 10:08:33] Welcome to Phoenix v2.0.0
[10:08:34] Couldn't connect to server, retrying...

Are you running it from a DOS windows using CMD?

This is what I'm typing in DOS window:

phoenix

This is what is in my cfg file:

backend = http://xxxxxx.worker2:[email protected]:8332

This is what I get:

[date time] Welcome to Phoenix v2.0.0
[0 Khash/s] [0 Accepted] [0 Rejected] [DISCONNECTED]

Given I can mine from the DOS window using poclbm.exe I know there isn't a communication / firewall issue. I'm really stuck trying to get Phoenix to work.

There is no command line configuration for phoenix 2.0 anymore, it can only be configured from the phoenix.cfg file.

Just copy and paste whatever I posted in my message into phoenix.cfg file (make sure you changed the pool, user name and password to something like that: backend = http://user.1:[email protected]:8337 (inside config file)).

Run phoenix from the command line: phoenix.exe (with no parameters).

This is how it is suppose to look like:
C:\Users\John Dow\Programs\phoenix2.0>phoenix.exe
[10/25/2012 21:44:34] Welcome to Phoenix v2.0.0
[21:44:35] Connected to server
[21:44:35] Currently on block: 204983
[21:44:40] [Streams] Result fdaaa8bf ACCEPTED
[331.71 Mhash/s] [1 Accepted] [0 Rejected] [RPC (+LP)]


I'm getting 337Mhs with my 5830 card with the configuration I posted. You can also configure and monitor phoenix miner directly from the web (once you get it running).

Your 5870 should give you around 400Mh/s

I found phoenix 2.0 to be thoroughly confusing.  The UI was awful as well.  I was glad to stop using it when I sold all my 5870s.

M

I'm not sure I understand your directions. I edited the cfg file and tried it taking out the -u and also taking out the parametes after 8332. Tried using command line option too. Nothing works. I just get messages saying "does not exist". Can you provide an example? I read that the config as shown in the directions don't work with the 5870 but have no more details to go by.

I was trying another one and got poclbm to work using command line. Trying to see if I can find something faster than GUIMiner I am using now. But poclbm has same speed on my 5870 of ~270 MH/s slightly overclocked running from command line as running it from GUIMiner.

Thanks!

Phoenix 2.0 is using config file for the parameters, modify phoenix.cfg located in the same folder:
Just run it with no parameters.

[general]
    autodetect = +cl -cpu #Use autodetect for all OpenCL devices, except those which are CPUs.
    verbose = False #Enable verbose logging?
    backend = http://xxxxxx.worker1:[email protected]:8332
    backups = http://xxxxxx.worker1:[email protected]:8332/
    failback = 300 #Seconds between attempts to reconnect to primary backend when using backups. (0 to disable)
    queuesize = 2 #Target/maximum size of the queue
    queuedelay = 5 #Seconds before work expires to request more work
    statusinterval = 1 #Seconds between statusbar updates
    ratesamples = 120 #Number of samples to average for hashrate reporting
    logfile = C:\inetpub\wwwroot\phoenix\log.txt #Set this option to log to a file.
[web]
    disabled = False #Disable the RPC server?

[cl:0:0]
    autoconfigure = false #Automatically configure this device?
    kernel = phatk2 #The kernel to use for the device
    start_undetected = False #Do not start the kernel if the device is not present
    disabled = False #Disable this device?
    worksize = 256 #Work group size, tweaking this option may improve performance
    vectors = True #Use unit2 vectors? (enable this or vectors4, not both)
    vectors4 = False #Use uint4 vectors? (enable this or vectors, not both)
    vectors8 = False #Use uint4 vectors? (enable this or vectors, not both)
    bfi_int = True #Use BFI_INT instruction on ATI VLIW GPUs?
    goffset = True #Use OpenCL 1.1 global offset?
    fastloop = False #Use fast internal loop? (ideal for low aggression)
    aggression = 13 #Number of nonces to test per kernel execution (lower value = less desktop lag, higher value = higher hashrate)

[cl:1:1]
    autoconfigure = true #Automatically configure this device?


   

Not having luck with Phoenix. I get the following message:

"Error: -u does not exist."

Using same user.worker info and password that works with GUIMiner. I'm running a HD 5870 GPU.

Here is what I tried:

Phoenix.exe -u http://xxxxxx.worker1:[email protected]:8332/ -k phatk platform=0 device=0 VECTORS BFI_INT AGGRESSION=10 worksize=128 FASTLOOP=false

phoenix.exe -u http://xxxxxx.worker1:[email protected]:8332/ -k phatk2 DEVICE=0 VECTORS BFI_INT AGGRESSION=4

i still use phoenix 1.7.5 with phoenix rising v1.3 from 5-31-11

it'd have to be multiple instances for it to be 10 times per second

one computer with 3 cards running w/ wrong settings could possibly do it (assuming it's not deliberate and just some guy running x instances of phoenix all on the same card)

I think phoenix support is dead.

RIP.

M

Phoenix 2 is spamming the wrong password 10 times per second, just like Phoenix 1. Maybe you can fix that?

When I try to run phoenix I get "Couldn't connect to server, retrying..."

What can I do to fix this?

Here is my cfg minus user/pass


I have tried with another pool as well.  And they are working in guiminer.  I want to see if I get better results using just phoenix instead of using phoenix through guiminer.  I get 5-10mh/s less using phoenix through guiminer than poclbm through guiminer.

Actually I don't want to became a miner developer, it's over my time capabilities. Maintaining a fork of miners with stratum protocol is quite overhead, accepting a simple patch by the mainline would be much easier, obviously.

best bet is to just add the support for the features you'd like directly, it is open source after all.   I think the original devs on phoenix have found other things to interest them for now, no commits in many months. 

Hello,

any chance to add support for Stratum mining protocol?

http://mining.bitcoin.cz/stratum-mining

Two pool are already supporting it (me and BtcGuild) and maybe other will join us, because that improvement against "getwork" is significant. Pools supporting Stratum have X-Stratum header in getwork response, so it is easy to switch to Stratum protocol transparently, without changing credentials/host/port by the user.

There are three ways how to integrate this stuff:

a) Startup stratum proxy as a separate process which will listen on some localhost port for getwork requests. Current Stratum proxy can handle this pretty easily.

b) Reimplement stratum protocol into the miner. It isn't too much difficult if you've already some experience with handling block headers and hashing.

c) This is the most likely solution for Python miners and especially these running on Twisted. I can add the feature to the proxy, that it will be able to run as a library. Potential interface would be extremely easy and if you already knows Phoenix internals, integration should be a matter of few minutes. Pseudo code looks like this:


What do you think about this? I'm not familiar with the Phoenix, but if I'll provide native Twisted library with such API, are you going to integrate it?

The major improvement in all this stuff is that proxy is producing block headers locally, so getwork() responses are *really* *really* fast. Also the network layer is improved significantly, so you need only up to 10kB/minute of bandwidth even for 18ExaHash/s (10**18) rigs.

Can this miner work with proxies? I'm behind a firewall.