freenode #monero

3 Mar 2020

[XMR] Be excellent to each other and welcoming to newcomers || getmonero.org || Dev: #monero-dev || Pool related: #monero-pools || Price chat: #monero-markets || OTC: #monero-otc || People can & do log this channel || DON'T USE Changelly, Freewallet or Minergate: reddit.com/r/monero/wiki/avoid

• 02:27
  kiwi_97
  Guys could anybody of you submit 5-7 shares to my pool? I have to test sth. which can just be done by one of you. Thanks
• 02:36
  riceandb1ans
  So I'm using xmrig for minig and wanted to build it on another system, noticed the donate-level option today and that it defaults to 5%. I changed it to 1% and my hash rate was reduced by 99%, is that expected?
• 02:44
  nioc
  riceandb1ans: that should not affect the hashrate at all
• 02:45
  nioc
  there is also #monero-pools but this is a slow time there as most who hang out there are euro
• 02:47
  nioc
  don't know how but something else must have changed
• 02:51
  riceandb1ans
  nioc: literally I took one box that was mining and did a SIGINT, then started it again with donate-level specified, hash rate went from 700 to 4
• 02:51
  riceandb1ans
  Nothing else was changed.
• 02:52
  riceandb1ans
  I guess they really like their donations lol
• 02:52
  nioc
  all mine are set at 1% and I get the expected HR
• 02:53
  nioc
  disclaimer: I am ignorant enough to not know what a SIGINT is
• 02:53
  ndorf
  are you sure it actually exited after the SIGINT? maybe you have two instances running
• 02:55
  nioc
  riceandb1ans: ^^ oh goody you are lucky, somebody who knows something is here
• 02:56
  riceandb1ans
  just one process
• 02:56
  riceandb1ans
  nioc: SIGINT, is the default signal on posix systems when you ctrl-c
• 03:05
  ndorf
  do you have enough RAM free?
• 03:06
  riceandb1ans
  Isn't it CPU and not RAM bound?
• 03:07
  ndorf
  it's both. you need 2-3GB of RAM, if it swaps you can kiss hashrate goodbye
• 03:07
  ndorf
  2-3GB for the xmrig process, to be specific
• 03:08
  riceandb1ans
  It does look like it briefly started paging because of monerod picking up a new block
• 03:08
  ndorf
  ideally, you would use hugepages and 1GB pages, in which case it would be a bit more than 3GB. that makes for better hashrate, but not a 100x difference
• 03:09
  riceandb1ans
  --randomx-1gb-pages use 1GB hugepages for dataset (Linux only)
• 03:09
  riceandb1ans
  This is on FreeBSD
• 03:11
  riceandb1ans
  Fear not, armed with anejo tequila, a lowball glass, and enough frustration, I'll be looking at source in a few hours when I'm the only one awake in the house
• 03:11
  ndorf
  try stopping monerod, starting xmrig, then restarting monerod
• 03:13
  ndorf
  fwiw has its own large pages (superpages). not sure if xmrig can use them though
• 03:13
  riceandb1ans
  That seems to have made a big difference
• 03:13
  ndorf
  FreeBSD has*
• 03:13
  riceandb1ans
  I'll look into it tonight
• 03:14
  riceandb1ans
  Also need to update the box I'm on :\
• 03:14
  ndorf
  that makes like a 5% hashrate difference though
• 07:29
  oyr884569023
  !skam
• 07:29
  Quotes
  dats a skam? hmm hmm nonono
• 07:39
  Randolf
  Ha ha ha!
• 11:15
  btgjoseph
  heyo
• 11:15
  btgjoseph
  exciting times xD
• 11:16
  btgjoseph
  fooorward!