Anyone zapping from Amethyst via Alby Hub noticing zaps are slow?

I'm zapping less because of that

Reply to this note

Please Login to reply.

Discussion

It’s likely not because of Amethyst, and just the LN routing being slow from an independent node.

But it's been consistently slow lately

What backend do you use for AH?

That's a good point. It could be an issue with the backend node, and not with Alby.

If it's routing, wouldn't the better question be who are your peers and channel partners?

I use their service

Is it limited to zaps via NWC, or is this lag also apparent with direct sends and invoices payments?

Try this!

lnbc10n1pnhrm20dqqnp4qv0nh8rz4lkk97vnntz2hs5dcjlhpvvc4pqskd35u4nkgch4jvh6vpp5fc8wfza4a5hd6cmm9t9849rfz4wyu8qtkz5r44rw6g0f5p6383rssp5l4z5lpd3wgyvg7tndawk6h9ddzt7h7mgjkw3kwf9vtjjqhdugp2q9qyysgqcqpcxqyz5vqu4ppfx4qwwuegasclqzmag3dvs2gg5ztl8gayxqfa3pqpp8cacf9n4067y3rfk8v0cnh3u9m8jzwf2762e3hz3ejew8a5legyte2nusph35yg7

That was fine. It took a little under 10 seconds to complete from the moment I submitted the send.

Hmm, could be how the NWC is designed. It uses the relay and passes all requests through it

If it's an Alby NWC, it would be using Alby's NWC relay, right? I think we'd expect the problem to manifest across the Alby user base if the embedded relay was the issue. No?

True, but other clients do not block on zaps, and it’s done async

Maybe I'm not patient enough? Sometimes taking over 10 seconds.

I have zapped notes and kept scrolling, when I come back I don't see it tallied.

Might be something in Amethyst. Would be nice to have an optimistic UI and notify when failure

Yeah, I am thinking a lot about it. Maybe some form of zap outbox that the app keep retrying if it fails while also building a list of outgoing zaps the user can see if succeeded or failed.

I am just too afraid of retrying stuff automatically. Maybe if it fails it just marks as such and allow users to retry manually.. it's less interesting, but safer.

Yes, autoretry should be avoided, or it will open a door to hacks and abuse

I think your instincts are right on. An automatic process has the potential to go haywire. Providing an opportunity to manually retry is, I think, the smarter solution.

Maybe retry just once after a fixed amount of seconds. If it still fails, send to the zap outbox for the user to retry manually

I use Rizful for my Lightning node, connected to Amethyst via NWC, and zaps are almost instantaneous. The purple wheel of death spins around very quickly, and the zap indicator lights up almost immediately. I don't use Alby anymore, but when I did it was similarly fast. I think the only times when it lagged were due to issues on the receiving end. Are you running your own backend, or are you using Alby Cloud to host your node?

Yes. Mine have been slow but I’m down to 1 channel, maybe that’s why

I’m on Damus though

I am on Amethyst with Albyhub, it works but yeah takes 4-5 seconds.