Ripple CTO David Schwartz has addressed concerns over a recent XRP Ledger incident, which saw nodes halting or not making progress for a brief period on the network.
In reaction to the incident, Bill Morgan, an XRP enthusiast and attorney, sought clarification on what developers meant when they described XRPL as “halted” rather than “down,” comparing the recent incident to past issues experienced by Solana.
I think the distinction is whether or not it’s still producing valid ledgers. If so, you can argue that it’s not down. But I think whatever technical arguments you want to make, if you can’t know you successfully completed a payment irrevocably, the system isn’t very useful.
— David “JoelKatz” Schwartz (@JoelKatz) November 26, 2024
Schwartz responded to this question saying, “I think the distinction is whether or not it’s still producing valid ledgers. If so, you can argue that it’s not down. But I think whatever technical arguments you want to make if you can’t know you successfully completed a payment irrevocably, the system isn’t very useful.”
Timeline of events: What happened?
At 2:45 p.m. on Nov. 25, Wietse Wind, a prominent XRPL developer, issued a tweet indicating that the XRPL might have halted: “The XRPL is currently partially down.We see all Full History servers operated by us and others down, and Ripple’s cluster is also reporting having no current ledger.”
In a further update, Wind stated that node operators were reporting their nodes halting/crashing or not making forward progress: “It seems the ledger has stabilised, and nodes are coming back and picking up on a new consensus ledger.” He added, “It seems ledger 92346896-92347095 have been lost.”
In another update, Wind indicated reports of XRPL nodes going down regardless of their node type: full history, current ledger, pathfinding and submission, and added that investigations will be launched into the cause of the halt.
In a final update about 40 minutes after the initial report, Wind indicated that XRP Ledger seems to have recovered on its own, and that no ledger may have been lost: “Most nodes are back on a consensus ledger and closing ledgers again.The Full History nodes are syncing fully back, meaning no ledgers seem to have been lost :)Network seems to recover on its own.Full History nodes are still syncing back, other than that we’re pretty much back to normal.”
In response to the incident, Brad Chase, VP of Engineering at RippleX, said the team is investigating the root cause and working on a fix, promising to share more updates as soon as additional information is obtained.
In another update, Chase told XRPL node operators and validators to update their infrastructure to the latest 2.3.0 rippled release as soon as possible, adding that given the nature of the issue, more details on the root cause might come “after” the network adopts the fix to ensure safety.
Additional details on the specifics of the incident was not yet available at press time.
