-
gbenson
are there settings/compile options that make it possible to run a node with under 4G ram, or is that a minimum? OOM killer gets me if I run with less:(
-
moneromooo
It should be more than enough unless you're syncing historical blocks. If that's what you're doing, check --db-sync-mode, tell it to sync more often - it'll be slower though.
-
moneromooo
Also check monerod --help for interesting things. Like... amount of memory to download new blocks before they're added. Forget the name.
-
moneromooo
Fewer peers might also help.
-
gbenson
it is syncing, does it use less when it's up to date then?
-
moneromooo
Yes.
-
moneromooo
Syncing means downloading lots of blocks, placing them in a queue, then adding them, typically 20 at a time. Plus syncing the DB every, er, lots.
-
moneromooo
When you're just keeping up with new blocks, add a block at a time, sync after each block.
-
moneromooo
Syncing more often while syncing historical blcks will kill performance, which is already not great.
-
moneromooo
But play with it, and see what you can get away with.
-
moneromooo
Or just keep > 4 GB till you're synced if you can, then lower.
-
gbenson
yeah, I think I'll do that, it's over halfway so hopefully done in a few hours
-
gbenson
would've been done already if I'd given the VM enough ram from the start :D
-
gbenson
--db-sync-mode default is fast:async:250000000bytes, so that's presumably 250Mb I could lose some of
-
moneromooo
FWIW sync speed is much faster at start for various reasons. Mostly tx volume being a trickle for the first few years.
-
moneromooo
But the last few thousand blocks are much slower as more is checked.
-
gbenson
does the database size on disk grow faster towards the end? it's at 54% using 20G disk for a pruned node, which would mean ~40Gb total *if* it grows at the same rate, but I was expecing more like 70G for a pruned node
-
gbenson
I guess if the first blocks are smaller that'd explain why it syncs a good way in before the OOM killer gets it
-
gbenson
afaict not setting --block-sync-size means it'll do 20 blocks at a time, but I don't know how big a block is so... :)
-
moneromooo
A pruned DB is 60 ot 65 GB or so I think.
-
moneromooo
Most early blocks are empty. Most late blocks are not. But IIRC adds are made 200 blocks at once for the early chain.
-
gbenson
ah, is that what BLOCKS_SYNCHRONIZING_DEFAULT_COUNT_PRE_V4 is
-
gbenson
why did empty blocks get added? (sorry if this is an idiot question)
-
moneromooo
To keep track of difficulty.
-
moneromooo
And if you add a rule that says you can't add empty blocks, miners will start adding their own dummy tx, so no gain, and a loss from spam.
-
gbenson
I never did quite figure out difficulty, I should probably re-read the original bitcoin paper again but the math was over my head the last time I tried :/
-
parazyd
Hey, what is the reason here that xmrig closes the connection after receiving a job?
termbin.com/8e5d
-
parazyd
Context, I'm trying to write a mining proxy
-
sech1
"target":"0x2" doesn't look right
-
sech1
and "id" field is missing in the response
-
parazyd
sech1: The stratum doc says it's a notification, so there should be no ID:
github.com/xmrig/xmrig-proxy/blob/master/doc/STRATUM.md#job
-
parazyd
The same disconnect happens when I use for example "b88d0600" for "target"
-
sech1
Try to connect to xmrig-proxy and find the difference in what you send and what xmrig-proxy sends in response to "login" message
-
sech1
oh, you definitely forgot to send "seed_hash" in the job. RandomX requires it
-
sech1
and "height"
-
sech1
-
sech1
There's quite a different in what it sends
-
sech1
first of all, it's not "params" but "result"
-
parazyd
So it means that STRATUM.md is outdated
-
sech1
absolutely
-
parazyd
Thanks for the link
-
sech1
-
sech1
with the addition of "algo", "seed_hash" and "height" fields
-
parazyd
oh
-
parazyd
Yeah missed that one
-
sech1
It's outdated because it doesn't mention "algo" extensions, but it's mostly accurate
-
parazyd
Alright, will see where that gets me
-
parazyd
sech1: It's getting me _somewhere_ now:
termbin.com/25ji
-
parazyd
I'm trying this just on a local-only monerod testnet node
-
parazyd
AFAICT it means that seed_hash is invalid
-
parazyd
What should it be on height 1 though?
-
parazyd
Since getblocktemplate() gives me an empty string for seed_hash on height 1
-
sech1
RandomX was activated at later height.
-
sech1
1978433 for mainnet, 1308737 for testnet
-
parazyd
ahhh
-
m-relay
<rucknium:monero.social> selsta: Statistical analysis of PR 9023 completed:
monero-project/monero #9023#issuecomment-1802593848
-
m-relay
<rucknium:monero.social> Summary: LGTM
-
Psylocybin42
Hello, I'm forking cryptonote and am having difficulty with the gui wallet. I have the core compiled and working on ubuntu16 and win10, and the gui wallet working on ubuntu 16. But I am not able to get the gui wallet to compile on windows. I make the symbolic link and git submodule, and these are the commands I use to cmake and build:
-
Psylocybin42
cmake -G "Visual Studio 15 2017" -A Win32 .. -DBOOST_ROOT=B:/local/boost_1_55_0 -DBoost_ATOMIC_LIBRARY="B:\local\boost_1_55_00\libs\atomic" -DBoost_SYSTEM_LIBRARY="B:\local\boost_1_55_00\libs\system" -DBoost_FILESYSTEM_LIBRARY="B:\local\boost_1_55_00\libs\filesystem" -DBoost_THREAD_LIBRARY="B:\local\boost_1_55_00\libs\thread" -DBoost_DATE_TIME_LIBRARY="B:\local\boost_1_55_00\libs\date_time"
-
Psylocybin42
-DBoost_CHRONO_LIBRARY="B:\local\boost_1_55_00\libs\chrono" -DBoost_REGEX_LIBRARY="B:\local\boost_1_55_00\libs\regex" -DBoost_SERIALIZATION_LIBRARY="B:\local\boost_1_55_00\libs\serialization" -DBoost_PROGRAM_OPTIONS_LIBRARY="B:\local\boost_1_55_00\libs\program_options" -DCMAKE_PREFIX_PATH="B:\Qt\Qt5.9.9\5.9.9\winrt_x64_msvc2017\lib\cmake\Qt5Gui;B:\Qt\Qt5.9.9\5.9.9\winrt_x64_msvc2017\lib\
-
Psylocybin42
cmake\Qt5Widgets"
-
Psylocybin42
MSBuild CryptoNote.sln /p:Configuration=Release /p:Platform=Win32 /p:PlatformToolset=v140_xp /m
-
Psylocybin42
But the build fails with a function overload error:
-
Psylocybin42
"B:\Documents\Crypto\cryptonotewallet-master\build\thecrowncoin.sln" (default target) (1) ->
-
Psylocybin42
"B:\Documents\Crypto\cryptonotewallet-master\build\cryptonote.vcxproj.metaproj" (default target) (3) ->
-
Psylocybin42
"B:\Documents\Crypto\cryptonotewallet-master\build\cryptonote.vcxproj" (default target) (5) ->
-
Psylocybin42
(ClCompile target) ->
-
Psylocybin42
B:\local\boost_1_55_0\boost/multi_index/ordered_index.hpp(272): error C2668: 'boost::make_reverse_iterator': a
-
Psylocybin42
mbiguous call to overloaded function [B:\Documents\Crypto\cryptonotewallet-master\build\cryptonote.vcxproj]
-
Psylocybin42
B:\local\boost_1_55_0\boost/multi_index/ordered_index.hpp(273): error C2668: 'boost::make_reverse_iterator': a
-
Psylocybin42
mbiguous call to overloaded function [B:\Documents\Crypto\cryptonotewallet-master\build\cryptonote.vcxproj]
-
Psylocybin42
45 Warning(s)
-
Psylocybin42
2 Error(s)
-
Psylocybin42
if anyone has any help or suggestions for compiling the cryptonote wallet from
github.com/cryptonotefoundation/cryptonotewallet on windows 10, then I would really appreciate it