I've same issue, but I think the problem is we're in Asia, the nsec.app relay and nostr.band relay are really slow and unstable in our location.

Reply to this note

Please Login to reply.

Discussion

nsec.app is running over relay.nsec.app which is a normal strfry. It's actually quote slow for ephemeral events processing - it has internal latency of at least 300ms for delivery of new events to active SUBs, plan to switching to something more specialized. Aside from that, again, try the secure enclave thing in settings.

I got some bug, I cannot upload my key and the ping is too high

1500 ping is probably due to you being in Asia, typical ping is 600 (300 of which are strfry).

Key not found sounds like service worker was off when you clicked Upload. Could you please reload the page and try again? Need to figure out how to wake the SW up if it starts sleeping with an open tab.

This TEE stuff is great, you found the holy grail I think. Are you running a two core graviton instance over there? Looks like the cheapest one is c6g.large. Or are you on 4 core intel?

I'm on intel, need to check arm too one day.

We're testing a secure enclave on Alibaba Cloud in Singapore, ping is 50-100ms. Testing AWS in Sing next. (Alibaba's enclave is basically a copy-paste of AWS, but their networking for SEA is very good.)

Is 50-100 ping from inside the enclave? Does enclave have network access by default or do you have to set it up somehow?

As low as 35ms for a light request from a test machine to the primary VM in Singapore (1700km one way), through vsock to enclave, light test process in enclave returns result, result sent back out the vsock and then from the primary VM back to the test machine.