Can we run a testing node on the same server as a standard masternode? It should use a different port right?
Maybe we can do a piggyback deal that automatically runs the testing node with a standard masternode.
You can, but it's a really bad idea. The server can crash, things could get corrupted, it's just not safe, and you'll lose your income if the server goes down. This can happen easily during a memory leak, where your resources are stretched to the limit, crashing the system, or causing your service provider to turn off your server.
Happens all the time in testing.
Over-the-top, worst case scenario assuming you don't already run autoreboot/autostart scripts (would get you back up and running before you are out of queue so no loss of income) and quite frankly, have a shitty service provider. With separate profiles, I have yet to see the base case for file corruption and the not safe comment is just bullshit.
The larger point made by eduffield is that we need a testing environment that more closely resembles mainnet. V12 didn't have more than a few hundred nodes and a few more people piggybacking aren't going to get us to 3000 testnet nodes.
Test-DASH is free but servers cost money - even on t-net
TESTER's need to be paid also - no different from the DEVs
My point I was trying to make originally, is that maybe a testing daemon could be built into the normal masternode dashd. Effectively, using the same servers as the main net without any extra server expenses while simulating the same network environment. I realize this isn't as secure or could cause problems, but it would sure be innovative if it could be pulled off.
hummm - so...
and hear me out..
DASH turned a FORK into a SPORK
Are you now asking that DASH turn MAIN-net into a PHASE 2 t-net??
WHY - THAT's COMPLETELY CRAZY !!!!
but - - - if you used a "mechanism" on Main-net - using PHASE 2 tech - you could talk backwards, to previous versions, to maintain a greater network stability moving forward - while then phasing out previous versions to create the consensus....
FOR EXAMPLE:::
When moving from V11 to V12 -
Test-net happened....
but - now that we have 'Phase 2' built-in (still the example)
We can phase-in new CODE while also phase-out old CODE
You could look at it as a SPORK on steroids with phases in Versions and not just features
[sorry but that sounds like a HUGE re-write..... so many CODE lines - ugh]
Maybe I'm thinking of Protocol Bumps - in that case - not so sure it could work
right then - I'm bat-shit crazy right now so - pay no attention to me - - - off to bed
ugh
either way - I'm all for it
[I'm thinking - Fractals]
edit: I still want to see multi-port MN's on the same IP - but - with a limit of (2-4) MN's per IP - I still think it could happen
How about calling it a rolling spork? Each node has a main port and test port. V12 is stable on main port. Install V13 on test port. Revise until stable. Then roll V13 to main port. Install V14 on test port. If the test port crashes - no biggie roll back to main port until new test version has a fix.
Or better yet, have the masternode figure out the new code to fix itself.....ok done watching Terminator. No Skynet.