u.today
The XRP Ledger (XRPL) is set to receive an upgrade aimed at resolving a network issue. The primary issue that prompted this upgrade involved full history (FH) nodes failing due to an SQLite restriction in page sizing.
Vet, an XRP dUNL validator, shared this development in an X post. According to Vet, the fix for the XRP Ledger Full History Servers issue has been merged in an official rippled release of the XRPLF Repository - rippled 2.2.3, which is now available for installation.
The fix for #XRP Ledger Full History Servers is merged in an official rippled release of the XRPLF Repository - rippled 2.2.3 and available for installation. pic.twitter.com/Owpd6xPEHx
— Vet 🏴☠️ (@Vet_X0) September 15, 2024
According to a screenshot shared by Vet, the Rippled (XRP ledger server) version 2.2.3 release is strongly recommended for full history servers using a page size of 4096. Since validators are not affected by this issue, they can either choose to run 2.2.2 or update to 2.2.3. This is because version 2.2.3 does not introduce any new amendments to version 2.2.2.
What happened?
Over the weekend, the XRP community drew attention to an issue on the rippled server that caused full history (FH) nodes to fail.
Under normal circumstances, full history servers happily record and serve up the full transaction history. However, FH nodes failed due to a SQLite restriction in page sizing.
This issue, according to some XRPL dUNL validators, had been highlighted weeks back but had failed to secure prompt attention.
XRP Cafe founder xrpl Adam clarified in an X post that, contrary to popular belief, this issue has no impact on consensus or network health.
Furthermore, with clio redundancy, most public XRPL endpoints do not require an actual FH server to return historical transaction results.
According to Xrpl Adam, the issue was not new on the XRP Ledger, having been documented years ago. It was noted then that certain rippled servers with full ledger history may experience an issue with their SQLite database page size, preventing the server from operating properly.
The XRP Cafe founder, however, believes that the urgency of the fix should have been promoted sooner to avoid FH nodes going down, which happened over the weekend.