Testnet is facing a problem with many pools being fully staked with millions of ADA, but abandoned and shut down, without any proper deregistration from blockchain. In result there is thousands skipped blocks each testnet epoch resulting in annoying slowness of whole testnet.
Right now TX confirmations already take forever due to relatively low amount of active working BPs (compared to mainnet).
Is there any plan to solve this? If this trend continues testnet will become unusable.
I’m committed to keeping a testnet pool operating for testing and QA purposes, ticker CHG if that helps you in any way.
not exactly, the problem here is that many people in the past were running pools there as well, just for testing, sometimes loaded with lots of tADA (high stake) and then when they lost interest they simply abandoned them.
Since on testnet there is no financial incentive to “sell” tADA to someone who cares or properly retire the pool (to get the 500 tADA back) they simply shut them down and let them registered to blockchain with all that huge stake left in their wallets.
That means these pools are still getting hundreds blocks assigned each epoch but they never mint them because they are offline, resulting in gaps in block production and overall slowness of the network.
Perhaps, “the community” – including wallet apps and everything – should decide to just create a new testnet every year and abandon it after two years.
By default, the testnet faucet Faucet currently doles out 1000 tADA per day, per developer, which is not close to an amount required for a stake pool to begin being elected to mint blocks. The faucet currently requests an API key to receive more tADA presumably, with no instructions on how to request an API key. Therefore, accumulating enough tADA to mint blocks requires some time and/or support from the developer community that may have access to more tADA than the faucet provides.
When a new testnet is created, how would potential SPOs be consistently supported to obtain enough tADA to continue minting new blocks without having to start from scratch to accumulate a sufficient amount to mint blocks?
SEAL, how do you monitor, observe or quantify the gaps in block production and slowness on testnet that you describe?
The faucet could be made better for such testnets regularly started from zero, with a possibility to automatically get an API key for the newly started one every year etc.
And the total stake would also restart from zero (or an initial batch could be distributed among the active users of the previous testnet). So, it would also not need so many tADA to consistently produce blocks.
Those requirements sound good to me.