Bitcoin Forum
December 08, 2016, 10:09:13 AM *
News: Latest stable version of Bitcoin Core: 0.13.1  [Torrent].
 
   Home   Help Search Donate Login Register  
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 [14] 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 »
  Print  
Author Topic: Pushpool - Tech Support  (Read 126760 times)
AndyRossy
Sr. Member
****
Offline Offline

Activity: 448


View Profile
July 03, 2011, 05:06:10 PM
 #261

Seems even wierder, as, first time i connect to it i do "get /" and it gives me back the 400 bad req, then, going forward, it just hangs.
1481191753
Hero Member
*
Offline Offline

Posts: 1481191753

View Profile Personal Message (Offline)

Ignore
1481191753
Reply with quote  #2

1481191753
Report to moderator
1481191753
Hero Member
*
Offline Offline

Posts: 1481191753

View Profile Personal Message (Offline)

Ignore
1481191753
Reply with quote  #2

1481191753
Report to moderator
1481191753
Hero Member
*
Offline Offline

Posts: 1481191753

View Profile Personal Message (Offline)

Ignore
1481191753
Reply with quote  #2

1481191753
Report to moderator
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
1481191753
Hero Member
*
Offline Offline

Posts: 1481191753

View Profile Personal Message (Offline)

Ignore
1481191753
Reply with quote  #2

1481191753
Report to moderator
1481191753
Hero Member
*
Offline Offline

Posts: 1481191753

View Profile Personal Message (Offline)

Ignore
1481191753
Reply with quote  #2

1481191753
Report to moderator
1481191753
Hero Member
*
Offline Offline

Posts: 1481191753

View Profile Personal Message (Offline)

Ignore
1481191753
Reply with quote  #2

1481191753
Report to moderator
AndyRossy
Sr. Member
****
Offline Offline

Activity: 448


View Profile
July 03, 2011, 05:59:10 PM
 #262

Last bitta info before anyone can help, i've got a little further, but now completely stumped.  It's a memcache issue by the looks of things.

It works fine w/out memcached

If I run memcached as current user, with "memcached -u pushpool -p 10000", and configure pushpool accordingly, it hangs. Same if I configure it to use the "servers" one which runs as demon on port 11211.

memcached output: (ran with -vv)

slab class   2: chunk size    120 perslab  8738
slab class   3: chunk size    152 perslab  6898
slab class   4: chunk size    192 perslab  5461
slab class   5: chunk size    240 perslab  4369
slab class   6: chunk size    304 perslab  3449
slab class   7: chunk size    384 perslab  2730
slab class   8: chunk size    480 perslab  2184
slab class   9: chunk size    600 perslab  1747
slab class  10: chunk size    752 perslab  1394
slab class  11: chunk size    944 perslab  1110
slab class  12: chunk size   1184 perslab   885
slab class  13: chunk size   1480 perslab   708
slab class  14: chunk size   1856 perslab   564
slab class  15: chunk size   2320 perslab   451
slab class  16: chunk size   2904 perslab   361
slab class  17: chunk size   3632 perslab   288
slab class  18: chunk size   4544 perslab   230
slab class  19: chunk size   5680 perslab   184
slab class  20: chunk size   7104 perslab   147
slab class  21: chunk size   8880 perslab   118
slab class  22: chunk size  11104 perslab    94
slab class  23: chunk size  13880 perslab    75
slab class  24: chunk size  17352 perslab    60
slab class  25: chunk size  21696 perslab    48
slab class  26: chunk size  27120 perslab    38
slab class  27: chunk size  33904 perslab    30
slab class  28: chunk size  42384 perslab    24
slab class  29: chunk size  52984 perslab    19
slab class  30: chunk size  66232 perslab    15
slab class  31: chunk size  82792 perslab    12
slab class  32: chunk size 103496 perslab    10
slab class  33: chunk size 129376 perslab     8
slab class  34: chunk size 161720 perslab     6
slab class  35: chunk size 202152 perslab     5
slab class  36: chunk size 252696 perslab     4
slab class  37: chunk size 315872 perslab     3
slab class  38: chunk size 394840 perslab     2
slab class  39: chunk size 493552 perslab     2
<3 server listening
<7 new client connection

--And now everything hangs, nothing sent back to miner or client at this point--

--If I kill memcached this is the output--

^CSIGINT handled.
pushpoold: libmemcached/io.cc:356: memcached_return_t memcached_io_read(memcached_server_st*, void*, size_t, ssize_t*): Assertion `0' failed.
pushpool@kokade:~/pushpool/sbin$

And then, pushpoold also falls over.

So yeah, that's what i've nailed it down to, but, cant seem to get any further? I dont like the idea of running this without memcached :S


d3c0n808
Member
**
Offline Offline

Activity: 113

Not Goxed


View Profile
July 03, 2011, 06:22:30 PM
 #263

Are you running the daemon on the same machine?  What distro is this?   What version of memcached are you using as well?  I know debian when you install memcached it starts up at boot so no need to invoke it by memcached -u (username).   Depending on the version you are using you could build a new package.  If pushpool is running it leads me to believe this is either a database issue or and rpc connection error.   Im using memcached 1.4.5

60 GH/s BFL Single SC - Pre-Order Yours Today!
`````` Only $1299.99 - butterflylabs.com ``````
AndyRossy
Sr. Member
****
Offline Offline

Activity: 448


View Profile
July 03, 2011, 06:55:15 PM
 #264

Are you running the daemon on the same machine?  What distro is this?   What version of memcached are you using as well?  I know debian when you install memcached it starts up at boot so no need to invoke it by memcached -u (username).   Depending on the version you are using you could build a new package.  If pushpool is running it leads me to believe this is either a database issue or and rpc connection error.   Im using memcached 1.4.5

Well, with memcached commented out - it runs + authenticates.

I am on Debian lenny,

Linux version 2.6.26-2-amd64 (Debian 2.6.26-26lenny2) (dannf@debian.org) (gcc version 4.1.3 20080704 (prerelease) (Debian 4.1.2-25)) #1 SMP Tue Jan 25 05:59:43 UTC 2011
pushpool@kokade:~/pushpool/sbin$ memcached -h
memcached 1.2.2

Any ideas? Really want memcached running.

And a new, second problem (when running without memcached), and a miner submits a share, I get this error.
 pg_sharelog failed: ERROR:  syntax error at or near ","
LINE 1: ..., reason, solution) VALUES (nextval('share_seq'),?,?,?,?,?,?...

Settings:
Quote
"SELECT password  from worker where account_username || '.' || password = $1",

#"stmt.sharelog":"insert into share (id,ourresult,reason,remotehost,solution, updatestreamresult, account_username) VALUES (nextval('share_seq'),?,?,?,?,?,?)"

"stmt.sharelog":"insert into share (id,rem_host, username, our_result, upstream_result, reason, solution) VALUES (nextval('share_seq'),?,?,?,?,?,?)"
},

If I connect via psql on command line, and run
insert into share (id,ourresult,reason,remotehost,solution, updatestreamresult, account_username) VALUES (nextval('share_seq'),NULL,NULL,NULL,NULL,NULL,NULL)
it works fine - has something changed in regards to the placeholders for sharelog since page #1 on this thread was made?

d3c0n808
Member
**
Offline Offline

Activity: 113

Not Goxed


View Profile
July 03, 2011, 07:01:07 PM
 #265

I would apt-remove memcached, and see if a custom built memcached works.  I had trouble with my setup and using the old libs and binary.  Although Im on kernel 2.6.37.6  I don't believe any placeholder has been changed ive been able to compile and run pushpool since 0.4 without issue.   You can get the src for memcached at http://memcached.org/  though if it compiled it may not end up being a memcached issue. 

60 GH/s BFL Single SC - Pre-Order Yours Today!
`````` Only $1299.99 - butterflylabs.com ``````
AndyRossy
Sr. Member
****
Offline Offline

Activity: 448


View Profile
July 03, 2011, 07:21:53 PM
 #266

Alright SQL problem fixed now - didnt realize postgres has it's own way of placeholding. 

Going to reinstall memcached shortly Sad but, all just seems wierd - there no way to get more verbose logging out of pushpoold as to why it's not liking memcached? What verison of memcached does it definatly work OK with?
d3c0n808
Member
**
Offline Offline

Activity: 113

Not Goxed


View Profile
July 03, 2011, 07:52:37 PM
 #267

No unfortunely the ammount of debbugging information provided by pushpoold is very very limited.   In my experience with pushpool.  I wasn't able to get it to compile without using the latest version of memcached.  But thats slackware not Debian.  I was able to compile on debian squeeze.  But I imagine the libs and binaries for lenny are older than for squeeze.    Though if pushpool can't connect to memcached it usually just exitted on my system.  

60 GH/s BFL Single SC - Pre-Order Yours Today!
`````` Only $1299.99 - butterflylabs.com ``````
Bullethead21
Jr. Member
*
Offline Offline

Activity: 43



View Profile
July 04, 2011, 03:11:45 PM
 #268

Stupid question time plz......can someone tell me how to properly shut down pushpoold please? I cant seem to find the command documented anywhere? Also, is there a clear example of how one's startup line should look to enable pushpoold to output display to the open terminal window?

like: ./pushpoold -F -E

This command seems to sorta work I think....all I get with that is the port list that pushpoold is listening on and that it is initialized..I dont however get output like for when a remote client connects?

Thanks in advance!

Xenland
Legendary
*
Offline Offline

Activity: 980


I'm not just any shaman, I'm a Sha256man


View Profile
July 04, 2011, 03:46:08 PM
 #269

Stupid question time plz......can someone tell me how to properly shut down pushpoold please? I cant seem to find the command documented anywhere? Also, is there a clear example of how one's startup line should look to enable pushpoold to output display to the open terminal window?

like: ./pushpoold -F -E

This command seems to sorta work I think....all I get with that is the port list that pushpoold is listening on and that it is initialized..I dont however get output like for when a remote client connects?

Thanks in advance!

d3c0n808, is correct pushpool is a cut and dry application it is very shy about giving details about what is happening although the command that you are using should display when users are connecting. I should note that it will only display when users are connecting on 2 of the 4 ports set in the .json file. One of the 2 ports I mentioned just now will work and the other wont, although they both report they are connected/connecting
froggy
Full Member
***
Offline Offline

Activity: 127


View Profile WWW
July 04, 2011, 07:47:42 PM
 #270

Quote
can someone tell me how to properly shut down pushpoold please?
I usually just..
Code:
ps a aux | gep pushpoold
then kill -9 puspoold's process number.

Brutal but effective.
   
d3c0n808
Member
**
Offline Offline

Activity: 113

Not Goxed


View Profile
July 04, 2011, 08:30:06 PM
 #271

Quote
can someone tell me how to properly shut down pushpoold please?
I usually just..
Code:
ps a aux | gep pushpoold
then kill -9 puspoold's process number.

Brutal but effective.
   


sudo killall pushpoold also works....

60 GH/s BFL Single SC - Pre-Order Yours Today!
`````` Only $1299.99 - butterflylabs.com ``````
Bullethead21
Jr. Member
*
Offline Offline

Activity: 43



View Profile
July 04, 2011, 11:49:57 PM
 #272

Awesome! Thanks guys!

I think I like the "sudo killall pushpoold" one the best though all methods dsescribed do indeed work.

Thanks again!

gigabytecoin
Sr. Member
****
Offline Offline

Activity: 280


View Profile
July 05, 2011, 02:15:30 AM
 #273

QUESTION: is /tmp/shares.log and /tmp/request.log entirely necessary? Under what circumstances would one require those files? I assume they would become quite massive in a short amount of time.

Send SIGHUP to the process, to re-open the logs (such as after rotation or deletion).



Why even create them at all though to begin with I guess is my question? If you are already making exact copies of them in your mysql database that you can clean by "timestamp" much easier?

If I simply remove those two lines from the server.json file, is everything going to explode?

Can anybody answer the above question?

Is it necessary to create the two log files when we are already backing up everything to MySQL?

Can I easily remove the two log file lines (or send them to a blackhole type script or a file that never saves or something?)

I plan on removing old "shares" from my MySQL table on the fly as a new block is found.
Xenland
Legendary
*
Offline Offline

Activity: 980


I'm not just any shaman, I'm a Sha256man


View Profile
July 05, 2011, 03:10:03 AM
 #274

QUESTION: is /tmp/shares.log and /tmp/request.log entirely necessary? Under what circumstances would one require those files? I assume they would become quite massive in a short amount of time.

Send SIGHUP to the process, to re-open the logs (such as after rotation or deletion).



Why even create them at all though to begin with I guess is my question? If you are already making exact copies of them in your mysql database that you can clean by "timestamp" much easier?

If I simply remove those two lines from the server.json file, is everything going to explode?

Can anybody answer the above question?

Is it necessary to create the two log files when we are already backing up everything to MySQL?

Can I easily remove the two log file lines (or send them to a blackhole type script or a file that never saves or something?)

I plan on removing old "shares" from my MySQL table on the fly as a new block is found.

Yes and no, The log files for for those who don't do terminal based debugging so I believe(don't quote the following) they are hard coded into pushpool to output them in case of a crash.
I think that log files usually trim the older lines of logs. Most developers(especially anyone that can program a pool servicing) has probubly been around for a while long before the days of when 512mb was unheard of and text data could file a whole hard drive within minutes. So I personally believe that they do stop getting bigger at some point

Can you easlity remove the two log file lines... Idk, i don't think anybody has tried, you should be the first. Let us know what your results are and we'll assist you wit what we can.

You should really install Mining Farm and check out the database structure to see how it all works I do something similar. Basically current rounds go to the `shares` table then upon a found block they get converted to the `shares_history` where they wait for the block to be confirmed, changing tables ensures better performance when doing reward counting. After its done rewarding and counting shares it moves all shares for that round into the `shares_dead` which is like a back up for operators in the case that the code went horribly wrong(hopefully never) and everyone got offset rewards. the operator can realy on a "backup" sorta of speak other then that `shares_dead` is just a back up Cheesy
gigabytecoin
Sr. Member
****
Offline Offline

Activity: 280


View Profile
July 05, 2011, 03:33:59 AM
 #275

Thanks for that reply Xenland.

And yes you are right, I should just get on to trying it and let everybody know! I will do so in the next few hours and report back on this thread as to whether or not you can safely comment out the two .log files specified in the server.json file that pushpool writes to.

(I was just hoping somebody had already attempted it!)

So what you are saying is you are not sure whether or not the .log files are automatically trimmed/kept in check by pushpoold?

If you are not doing anything with the .log files from mining farm and it has been used by a few pools now in actual production, I would assume that pushpool deals with them directly otherwise people's servers would probably be getting overloaded with them by now, or soon.

But why then would jgarzik mention how to "re-open the logs (such as after rotation or deletion)." if pushpool dealt with them directly?!

So confused!

I will try it out myself and report back ASAP.
jgarzik
Legendary
*
Offline Offline

Activity: 1470


View Profile
July 05, 2011, 03:59:31 AM
 #276

Sending a normal SIGTERM signal is the normal way to shut down pushpoold.  So killall works... provided that it does not send a SIGKILL (kill -9) immediately.

pushpoold will receive the SIGTERM, and initiate safe shutdown procedures, closing files and database connections properly, etc.


Jeff Garzik, bitcoin core dev team and BitPay engineer; opinions are my own, not my employer.
Donations / tip jar: 1BrufViLKnSWtuWGkryPsKsxonV2NQ7Tcj
Xenland
Legendary
*
Offline Offline

Activity: 980


I'm not just any shaman, I'm a Sha256man


View Profile
July 05, 2011, 05:33:43 AM
 #277

Yeah i dont know what happens in a scientifc proven method to the log files but i do alot of testing with pusgpool on my local server and the log files seem to stay in the same range from a first glance. Im pretty curious about this problem as well and will be sure to try it out after celebrating.

Happy 4th of July everyone.


Thanks for that reply Xenland.

And yes you are right, I should just get on to trying it and let everybody know! I will do so in the next few hours and report back on this thread as to whether or not you can safely comment out the two .log files specified in the server.json file that pushpool writes to.

(I was just hoping somebody had already attempted it!)

So what you are saying is you are not sure whether or not the .log files are automatically trimmed/kept in check by pushpoold?

If you are not doing anything with the .log files from mining farm and it has been used by a few pools now in actual production, I would assume that pushpool deals with them directly otherwise people's servers would probably be getting overloaded with them by now, or soon.

But why then would jgarzik mention how to "re-open the logs (such as after rotation or deletion)." if pushpool dealt with them directly?!

So confused!

I will try it out myself and report back ASAP.
Xenland
Legendary
*
Offline Offline

Activity: 980


I'm not just any shaman, I'm a Sha256man


View Profile
July 06, 2011, 02:41:02 PM
 #278

Just thought I'd update the install tut scince I have to reinstall the whole thing on my VPS for further testing of my Mining Farm software, might as well get this outta the way.

Thanks!
jgarzik
Legendary
*
Offline Offline

Activity: 1470


View Profile
July 06, 2011, 05:30:19 PM
 #279

If you are not doing anything with the .log files from mining farm and it has been used by a few pools now in actual production, I would assume that pushpool deals with them directly otherwise people's servers would probably be getting overloaded with them by now, or soon.

But why then would jgarzik mention how to "re-open the logs (such as after rotation or deletion)." if pushpool dealt with them directly?!

You must rotate/trim the logs yourself.

The logrotate package can help you with that.


Jeff Garzik, bitcoin core dev team and BitPay engineer; opinions are my own, not my employer.
Donations / tip jar: 1BrufViLKnSWtuWGkryPsKsxonV2NQ7Tcj
redshark1802
Jr. Member
*
Offline Offline

Activity: 44


View Profile
July 08, 2011, 11:02:21 AM
 #280

Hello,

i've played a bit aroung with pushpoold and everything worked just fine.
But when I asked a friend for a little test something strange happened.
I have about 100MH/s my friend has 1500MH/s.
As soon as my friend joined for testing i got a lot of stales,
I looked it up in the datase and the reason is "unknown-work".
Where does this come from?

Help with this would be really nice.


regards, redshark1802

If the post above is something worth to you, please consider donating Wink.
1HYek9jXvbMDYyYwN7rxy9i3oDWgCYhATx
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 [14] 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 »
  Print  
 
Jump to:  

Sponsored by , a Bitcoin-accepting VPN.
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!