13:03:25 <ruien> Hey everyone. I need to sync a stagenet node in userspace on limited RAM (<2GB) on a machine with 192 GB of RAM. I set 'ulimit -v 1572864' as it looks like DEFAULT_MAPSIZE is 1GB (lmdb/db_lmdb.h). Even syncing with '--max-concurrency 1' though results in allocation errors: 'Exception at [add_new_block], what=std::bad_alloc'. Is it just not possible to sync on this ram size, or is this something about the system resources being autodetect
13:03:25 <ruien> ed and ulimit not being respected?
13:09:50 <ruien>  /msg NickServ VERIFY REGISTER ruien Y8Cfz7COFPRL31SH
13:09:55 <ruien> oops
13:12:00 <Inge> ooops
13:16:56 <moneromooo> LMDB will need loads of virtual memory. Don't constrain it.
13:17:30 <moneromooo> At least as much as the db size, so... 128 GB or so ?
13:18:39 <ruien> That makes sense. Actually I don't want/need to constrain virtual memory; I really just want to constrain RSS. Something like "operate as if this machine had 2GB or 4GB instead of 192 GB". But maybe this is not feasible at all.
13:19:11 <moneromooo> Maybe look into cgroups.
13:19:47 <ruien> yes, I think cgroup v2 has something for this, but I was hoping to do it entirely in userspace. Guess not.
13:20:01 <ruien> Thanks for the response though - i appreciate it.
13:22:53 <moneromooo> There are options to tell lmdb to flush often. --db-sync-mode IIRC. It'll slow things down though.
13:23:08 <moneromooo> But it *should* cut down on RSS a bit I think. YMMV.
13:23:40 <ruien> Slow is no problem here, so I'll look into that as well.
13:24:25 <moneromooo> Running in docker/podman might also add some easy to use memory constraints.
13:28:10 <ruien> '--db-sync-mode fastest:sync:20blocks' is looking really good
16:03:38 <hyc> funny. I've never seen my own ping timeout before: (15:13:57) hyc left the room (quit: Ping timeout: 240 seconds).
16:05:58 <moneromooo> That happens when someone sees you ping out, quickly joins with your nick and says something outrageous quick before you come back.
16:09:31 <hyc> uhhhhh. ok. would def have to be quick
16:10:15 <moneromooo> So... you're not really planning on a patch to use leveldb in monerod then ?
16:10:53 <moneromooo> There was something in that which I found a bit odd, but I couldn't quite put my finger on it...
16:11:46 <hyc> gag.....
18:45:13 <saunders> where can I find raw historical data for monero v bitcoin prices?  I can google "price monero bitcoin" and get nice charts.  but where does the data originate?
18:46:12 <hyc> coingecko lets you download CSVs of their data
18:51:04 <apotheon> saunders: It must ultimately come from exchnage services.
18:51:16 <apotheon> s/hage/ange/
19:02:39 <noex> my monero node synced up to within like 100 blocks of being fully synced, but it's kind of stuck here. sometimes it's 70 blocks behind, then 100, then 80...anyone experience that?
19:03:11 <noex> it's been hanging out like this for a few days now
19:03:38 <noex> maybe that's normal, I don't know
19:18:23 <Mochi102> noex, try exiting the daemon and restarting it
19:59:42 <noex> Mochi102: is ctrl-c safe?
19:59:59 <noex> oh wait i detacted it
20:00:07 <noex> detached*
20:00:42 <noex> can i just kill it?
20:01:03 <noex> going to run it in tmux and not detach this time
20:01:20 <ndorf> SIGTERM should be safe, alternatively you can try `monerod exit`
20:01:34 <ndorf> (SIGTERM is the default when you kill with no specific signal)
20:01:39 <ndorf> don't kill -9
20:02:01 <noex> it wasn't happy about exit
20:02:09 <ndorf> what happened?
20:02:19 <noex> Error: Daemon did not stop-- rpc_request:
20:02:19 <ndorf> i just tried it and it worked as expected
20:02:36 <ndorf> interesting. maybe your default RPC port is restricted? does that ring a bell?
20:02:48 <noex> oh yeah, i did set that
20:02:54 <ndorf> that would do it
20:02:55 <noex> i'm trying to run a public node
20:03:02 <noex> and i have no idea what i'm doing lol
20:03:04 <ndorf> so then just kill <pid>, no other arguments
20:03:32 <noex> alright that worked
20:03:51 <noex> maybe i should share my config as a sanity check
20:04:27 <noex> i might be doing something really stupid
20:05:50 <noex> https://dpaste.com/5PYX5KD2C
20:07:52 <noex> i suppose i could also use a configuration file, at least I think...but I ran it like this
20:13:23 <apotheon> Wow, I screwed up my regex earlier.
20:14:11 <apotheon> I came back to this window *just* in time to notice.
20:17:11 <ndorf> noex: looks mostly ok
20:17:26 <ndorf> i think rate limits are in KB, so unless i'm mistaken yours are both set to 1GByte/sec
20:18:41 <noex> so the only reason I added that is because some page somewhere mentioned that the default limits were really low
20:18:49 <noex> but I guess I can just try running without that
20:18:49 <ndorf> a couple of the other options are superfluous, since the values are the same as the defaults
20:19:50 <noex> yeah, i was trying to be explicit...probably no point really though.
20:19:54 <ndorf> i think the defaults are 2048KB/s
20:19:55 <noex> the ports and bind ips i think are defaults
20:20:07 <ndorf> if you prefer to have them explicit, that's fine. just pointing it out
20:20:37 <ndorf> i'm not sure if --rpc-login makes sense in conjunction with --public-node and/or --restricted-rpc?
20:20:42 <noex> sure, i appreciate it. i might just get rid of them now that i kinda know what the ports are anyway
20:21:00 <noex> yeah that part confused me a bit
20:22:01 <ndorf> if you want anyone to be able to use your node's RPC, you wouldn't want --rpc-login
20:22:19 <noex> lol
20:22:24 <noex> that...makes a bit of sense
20:22:24 <ndorf> but if you don't, you wouldn't want to set --public-node at all, and probably not --restricted-rpc either since it's secured by the login
20:22:36 <noex> i definitely wanted a public node
20:23:21 <ndorf> also note that the bandwidth limits apply only to P2P, not RPC, last i checked
20:23:34 <ndorf> RPC is always unlimited, i think, and can use quite a lot depending on clients
20:23:58 <noex> i just got rid of those limits. some blog or something said to add it but I guess if it's not broken there's no sense in preemtively fixing it
20:25:52 <ndorf> looks like the defaults are 1MB/s, not 2 like i said earlier
20:26:09 <ndorf> but again, that only applies to P2P i'm pretty sure
20:28:01 <noex> the documentation in one place says CPU mining is fine and somewhere else says people use GPUs and not CPUs
20:29:39 <ndorf> GPUs are not useful for mining Monero, only (some) CPUs
20:30:42 <noex> that's what i thought
20:30:52 <ndorf> if something says use GPUs, it's probably outdated, from before 2019 (when a different PoW was used)
20:31:30 <noex> ohh.  yeah it says "Be advised though that real mining happens in pools and with high-end GPU-s instead of CPU-s."
20:32:25 <noex> i was hoping to CPU mine on another machine as well
20:33:52 <noex> is it actually not hopeless like every other crypto? lol
20:34:44 <ndorf> well, that depends on your CPU :)
20:35:11 <ndorf> my ryzen 3600 recently became unprofitable, like within the past couple of months
20:35:44 <ndorf> 3900X is still in the green i would imagine
20:35:57 <noex> i have a xeon w-11955m
20:37:11 <noex> --public-node is enabled, but RPC server 127.0.0.1:18081 may be unreachable from outside, please check RPC server bind address
20:37:17 <noex> ^ this message is confusing
20:38:15 <Mochi102> How is it confusing?
20:38:21 <noex> i don't get why it needs a bind address other than localhost
20:38:22 <ndorf> hm, i guess you need to specify --rpc-bind-ip
20:38:35 <ndorf> seems like it should be implied with --public-node
20:38:47 <ndorf> well, if it's only reachable from localhost, that's not very public now is it? :)
20:38:58 <noex> isn't that the purpose of port forwarding?
20:39:27 <ndorf> port forwarding from where? your router? 
20:39:35 <noex> yeah
20:39:44 <ndorf> that won't work unless your router is on the local host
20:40:01 <Mochi102> What if you have multiple IP addresses?
20:40:02 <ndorf> in which case... you don't need forwarding, heh
20:40:04 <apotheon> I don't recall my specific Ryzen CPU model.
20:40:16 <Mochi102> Probably want to bind to one.
20:40:24 <ndorf> if you have multiple IP addresses, you can pick one to bind to, or you can bind to 0.0.0.0 which means all of them
20:41:01 <Mochi102> ^^
20:41:14 <noex> all...local ips?
20:41:30 <apotheon> ndorf: Whether a 3900 is still profitable depends (in part) on the price of electricity where you live.
20:42:04 <ndorf> well yeah.
20:42:19 <apotheon> In Denmark, it's probably not profitable.
20:42:19 <noex> i don't get the difference between binding to the local ip and using 127.0.0.1
20:42:24 <apotheon> That's my guess, anyway.
20:42:33 <ndorf> if you bind to 127.0.0.1, it's only reachable from the same host.
20:42:44 <ndorf> your router won't be able to reach it.
20:43:07 <noex> it wouldn't be able to reach it anyway if i didn't forward the port to it
20:43:30 <ndorf> the router can't forward ports to it because it can't reach it
20:43:45 <noex> how can my router not reach a local machine on my network
20:43:54 <ndorf> because it's not listening on the network, it's only listening on localhost.
20:44:41 <noex> so i just have to hardcode the machine's IP address?
20:44:58 <ndorf> you can do that, or you can use 0.0.0.0 which means listen on all its IP addresses
20:44:59 <noex> or like ip link | grep | cut ?
20:45:58 <noex> okay
20:46:16 <noex> so by default it won't listen to anything but itself
20:47:21 <ndorf> correct
20:54:31 <noex> this is 100x better in interactive mode. i didn't realize it had a cli interface.
20:55:55 <ndorf> yeah, you can run that in tmux or something
20:56:52 <noex> it seems to block and awful lot of hosts
20:58:59 <ndorf> yup
20:59:06 <noex> why does that happen?
20:59:06 <ndorf> it's a dangerous world out there
20:59:16 <noex> just malicous stuff in the wild
20:59:37 <ndorf> yep
21:02:30 <asymptotically> ndorf: 3900X makes around £0.50 a day in current network conditions. think it costs me around £0.60 in electricity to run at full load for a day though :$
21:02:50 <ndorf> oof.
21:09:50 <noex> i might have killed it. i typed "check_blockchain_pruning" just experimenting and it's been frozen ever since.
21:14:11 <noex> the kernel is using a bunch of CPU but not monerod. what have I done?
21:26:23 <ndorf> i don't actually know what that command does but if it's reading the whole blockchain, that could explain it (kernel is doing the I/O)
21:26:51 <noex> it also is ignoring SIGTERM
21:27:11 <noex> i dug my grave, and now i will lay in it
21:34:21 <apotheon> s/lay/lie/
21:35:24 <apotheon> Lay should be used as a transitive verb.
21:35:33 <apotheon> When you need the intransitive, use lie.
21:44:58 <ndorf> i find it's easier to just lie all the time.
21:47:05 <noex> i dug my grave, now i'm going to lay in it and tell lies
21:49:01 <noex> "bitcoin cash is the real bitcoin"
21:50:10 <saunders> noex: you tell 'em
21:50:10 <noex> kidding. also i kind of wish it was at this point.
21:51:19 <noex> bitcoin is such a bizarre phenomenon
21:53:23 <noex> i've kind of given up on at at least for what it was designed for. i feel like monero is closer in line with the original vision.
21:54:41 <noex> ah pruning check done. it says it's all good. nice
21:54:56 <saunders> bitcoin is a deflationary asset while monero is inflationary.  don't ask me to explain that, because I don't quite know what means.  I can say that bitcoin is analogous to "e-gold" because of the finite number of coins.
21:56:31 <noex> i don't think that is what satoshi intended at all, but you are definitely right. i think the fact that it is slow/expensive to use compounds the problem.
21:56:54 <noex> like i no longer feel okay buying normal things with bitcoin
22:02:18 <noex> i don't know if other coins now level the playing field with the algorithm (ASIC-resistant) but i feel like that is definitely the right choice.
22:04:21 <apotheon> saunders: Whether Monero is inflationary or deflationary actually depends on how much of it will get lost.
22:05:10 <apotheon> noex: Monero basically solves all the biggest problems BTC has as a currency.
22:05:45 <apotheon> I don't recally refer to "Bitcoin" any longer.  Doing so means taking sides in the split.  I just refer to BTC as BTC, and BCH as BCH or Bitcoin Cash.
22:06:02 <apotheon> . . . though if I had to pick one to survive, it'd probably have to be BCH.
22:06:28 <noex> i thought that too. if you think about it logically, how can this be sustainable? it makes no sense.
22:07:00 <apotheon> I guess I sometimes refer to "Bitcoin" when what I mean is what Satoshi made, no longer than after the final Great Vanishing.
22:07:38 <apotheon> Yeah, BTC seems to be trying to make itself suck as a currency and "solve" the problem by coming up with excuses meant to sway people who stopped believing.
22:07:40 <noex> what satoshi wrote and what we have now are just so at odds with each other it's comical
22:07:44 <apotheon> LN isn't a solution: it's an excuse.
22:08:00 <apotheon> noex: Speak for yourself.  I'm using Monero.
22:08:05 <apotheon> That's pretty close.
22:08:33 <apotheon> Well, in theory I'm using Monero.  I seem to have lost mine on a recent oceanic vacation.
22:08:58 <noex> it is way closer and even improves on it imo. i don't think anyone had the foresight to predict something like Bitmain 
22:09:43 <apotheon> I've gotta go do some writing.  Ta.
22:11:11 <noex> looking at what crypto has become is crazy. dogecoin? it's just like a big meme at this point. my buddy bought a bunch of SHIB hoping to "get rich".
22:11:56 <noex> it's more like a lottery. i wish i had stumbled across monero sooner because it just seems like what we actually wanted in the first place.
23:48:47 <noex> https://dpaste.com/C337UAW4Z same problem. i'm falling further behind while syncing
23:53:39 <noex> it synced fine up to ~100 blocks behind but once I got here it's been stuck for a couple days