Hey guys, I'm wondering how subaddress scales.

Can monero handle 100,000+ subaddresses in one wallet?

ye

is that a yes?

it is

afaik more subadresses is basically like having a hash list to check against incoming tx, so very little overhead per extra subaddress

and you dont need to check each individual address for a payment, you can just look at the incoming transactions

it'll tell you which subaddress it's being paid into

yeah that one I know

does the gui/cli wallet make handling 100k addresses easy? i'd imagine not

you can label them

nah I'll be using rpc, no worries

guys, if double_spent_seen is true, what steps should be taken?

Typically increase the number of confirmations you require before assuming you'll keep the money long term.

I'm also curious how the daemon will react if a block is orphaned

if I get_transfer and get tx in an orphaned block, how does the daemon notify me?

It doesn't. It has no idea what txes are yours.

ah

Well, scratch that. I'll assume you mean the wallet daemon. I'm not sure exactly, but some combination of --tx-notify (in the wallet) and --block-notify (in the node) should get you that.

I'm trying to judge the best way.

I need to receive payments.

So, at first I thought I'll use tx_notify to watch for payments

but then I need to track confirmations of indiv txes.

I think using block-notify, and then sorting out txes with confirm > 3 is the best way

waddya guys think?

From memory, when I did this, I used get_bulk_transfers, and triggered at >= N confirmations.

I think you are referring to get_bulk_payments but payment_id is phased out so..

Quite possibly, the name was from memory.

If you give an empty list of payment ids, it returns all.

After computer has been suspended and woke up, I had monerod using a whole core fulltime even after bieng fully synced. Logs where non stop this:

2019-12-14 11:45:03.687 T Throttle throttle_speed_in: packet of ~0b (from 0 b) Speed AVG= 0[w=9.205] 0[w=9.205] / Limit=16 KiB/sec [0 0 0 0 0 0 0 0 0 0 ]

2019-12-14 11:45:03.687 T Throttle <<< global-IN: packet of ~0b (from 0 b) Speed AVG= 2[w=9.205] 2[w=9.205] / Limit=8192 KiB/sec [0 0 223 0 0 0 223 2655 5310 10620 ]

2019-12-14 11:45:03.687 T Setting -13:42:10.984143 expiry

2019-12-14 11:45:03.687 E Setting timer on a shut down object

I guess the negative timer has to do with the suspending. No other clue what exactly was using one core non stop in the logs.

any chance the monerod binary tests for AVX2 instruction support during build?

set_log to 0 till that's gone

I'm trying to run a binary I built on one computer on another computer but it segfaults with illegal instruction

The throttle logs are *very* noisy when there's been a long pause.

the instruction being andn

Hmm. Nevermind. Negative seems bad, yes.

I don't mind about the logs, but it was using a lot of CPU. Gave it 20mn and then restarted it, then all good.

Then build it on that computer, or set the arch to something the second computer supports.

I think I had that twice, but did not care first time as I had to reboot shortly anyway. So I can probably reproduce.

moneromooo: how do I set the arch for the monero build

The -march or -cpu option.

One is old, I never remember which.

binaryFate: how much time did that computer sleep ?

(roughly)

this is what I have right now git.archlinux.org/svntogit/communit…runk/PKGBUILD?h=packages/monero#n44

night + morning, The -13h timer might be correct

hey guys, I'm getting problematic log lines

nvm, figured ouit

kpcyrd: if you look at CMakeLists.txt, it uses both -march and -mcpu. Use those. I think I was thinking of -mtune for the obsolete one.

IIRC -march is what you want and -mcpu is the tune replacement.

So you'd use -march=i386 to ensure the code gen can work on all i386 and later.

The GCC doc likely lists the available archs.

I'd just try -D ARCH="x86-64".