- A failure of the Ethereum 2.0 Spadina testnet will require the launch of a second test, Zinken.
- Prysmatic Labs’ report rules out critical error in testnet and attributes it to fixable conditions.
The coordinator of the Ethereum 2.0, Danny Ryan, has confirmed a failure on tesnet Spadina. Designed to be “dress rehearsal” before the launch of phase 0, the testnet was deployed on September 29th with 2,856 active validators and 91,382 ETH staked. At 19:35 (UTC) Ryan posted that Spadina had been justified.
The coordinator of the Ethereum 2.0 stated that the test network showed several problems in its configuration, the client nodes, in calculation errors genesis. The failures were exacerbated by the low participation in Spadina along with “small failures” in the process of launching the clients. Low participation for Ethereum 2.0 has been a constant concern of the Ethereum core developers and the community.
Details on Ethereum 2.0 testnet failure
Ryan announced the launch of a second “dress rehearsal”. Therefore, next week they will deploy the Zinken testnet. However, for this testnet Ryan will focus on a “clean client release process” and a better user experience. Ryan added:
As this is a dress rehearsal, we ask you to take genesis seriously. Only make deposits for vals you intend to run, and if at all possible, be attentive in the 24 hours leading to genesis — upgrading your node if necessary.
The main bug seems to have originated, according to Ryan, on the client Prysm. The Ethereum client has published a report outlining the sequence of events and causes that led to the failure. According to the Prysmatics Labs report, at 1:35 p.m. (UTC) the team identified a bug and began an investigation. It was later determined that the client lacked the necessary bootnodes and that Spadina had not been added to the validator commands for the Prysm alpha.26 version. Therefore, Prysm’s deposits were invalidated. The report points out:
(…) our validator commands did not have the –spadina flag, leading to invalid deposits for anyone who created a wallet with Prysm and tried to deposit using the deposit data from their created accounts. Additionally, our documentation portal had 0 mention of Spadina. We lagged behind in terms of taking Spadina seriously, having a detailed checklist, and overall having a release ready before the genesis event.
Prysmatics Labs indicates that the bug caused confusion amongst the users, no finality on the testnet, low participation, in conjunction with invalid deposits. The report states that the bug gave a “bad look to the genesis rehearsal”. However, Prysmatics determined that the bug was not critical or related to a consensus error. The Prysmatics Labs team has deployed an update for its node (alpha.27) and has informed the community and its users through its social networks.
In mid-August this year, Medalla testnet also experienced a bug related to a drop in participation. The failure caused the appearance of unsynchronized nodes. Like the Spadina error, only Prysm was affected. Since then, the bug has been fixed. For now, the launch of the highly anticipated phase 0 is still scheduled for November 2020.