SDL selects a server that is down during wallet creation/restore #135

getRandomServer
Checking if lite server  "https://lite.hushpool.is"  is a alive, try= 0
Checking if lite server  "https://lite2.hushpool.is"  is a alive, try= 1
Checking if lite server  "https://lite.hush.is"  is a alive, try= 2
Checking if lite server  "https://lite.hush.community"  is a alive, try= 3
Checking if lite server  "https://lite.myhush.org"  is a alive, try= 4
Checking if lite server  "https://wtfistheinternet.hush.is"  is a alive, try= 5
Checking if lite server  "https://poop.granitefone.me"  is a alive, try= 6
Checking if lite server  "https://lite.hush.land"  is a alive, try= 7
Lite server in conf  "https://lite.hushpool.is"  is down, getting a random one
getRandomServer
Checking if lite server  "https://poop.granitefone.me"  is a alive, try= 0
Checking if lite server  "https://lite.hush.land"  is a alive, try= 1
Checking if lite server  "https://lite.hushpool.is"  is a alive, try= 2
Checking if lite server  "https://lite2.hushpool.is"  is a alive, try= 3
Checking if lite server  "https://lite.hush.is"  is a alive, try= 4
Checking if lite server  "https://lite.hush.community"  is a alive, try= 5
Checking if lite server  "https://lite.myhush.org"  is a alive, try= 6
Checking if lite server  "https://wtfistheinternet.hush.is"  is a alive, try= 7
getSettings
"loadConnection" :  calling doAutoConnect
getSettings
"doAutoConnect" :  Creating connection with server= "https://poop.granitefone.me"
"doAutoConnect" :  Attempting to initialize library with  "https://poop.granitefone.me"
Wallet exists: false
"doAutoConnect" :  no existing wallet
createOrRestore : server= "https://poop.granitefone.me"
FirstTimeWizard : dangerous= false  server= "https://poop.granitefone.me"
FirstTimeWizard : backed up old addresslabels

It says lite.hushpool.is" is down, getting a random one, which is incorrect.
SDL says poop.granitefone.me is a live, which is in fact was down for some time.

image.png

48 KiB

It's getting more interesting, no matter what it now selects the same server:

[connection]
server=https://poop.granitefone.me

I am pretty sure I have seen this in the past, i.e. this is not a new bug in 2.0.0. Also, getting the "unexpected compression flag" error doesn't mean the server is down, it means the TLS connection has been corrupted/broken.

@onryo a lot of stuff has changed recently on the dev branch related to this bug/error. Can you confirm if this still happens on the latest dev branch?

@onryo my guess is that the latest PR into the dev branch has fixed this

I confirm it's fixed.