Got a response from jbellis in IRC saying that the node will have to build its own hash tree. The request to itself is normal.
On Mon, Sep 19, 2011 at 7:01 AM, Jason Harvey <alie...@reddit.com> wrote: > I have a node in my 0.8.5 ring that I'm attempting to repair. I sent > it the repair command and let it run for a few hours. After checking > the logs it didn't appear to have repaired at all. This was the last > repair-related thing in the logs: > > INFO [AntiEntropyStage:1] 2011-09-19 05:53:55,823 > AntiEntropyService.java (line 215) Sending AEService tree for > #<TreeRequest manual-repair-843e8719-f3b2-4607-8eda-1cd766248c26, > /10.40.50.89, (reddit,SavesByAccount), > (68056473384187692692674921486353642290,102084710076281539039012382229530463435]> > INFO [AntiEntropyStage:1] 2011-09-19 05:53:58,898 > AntiEntropyService.java (line 215) Sending AEService tree for > #<TreeRequest manual-repair-843e8719-f3b2-4607-8eda-1cd766248c26, > /10.40.50.89, (reddit,SubredditsByPartialName), > (68056473384187692692674921486353642290,102084710076281539039012382229530463435]> > INFO [AntiEntropyStage:1] 2011-09-19 06:32:44,750 > AntiEntropyService.java (line 215) Sending AEService tree for > #<TreeRequest manual-repair-ff545576-a71d-49e4-9a3d-b58af94c5d99, > cass-40/10.40.39.113, (reddit,CollegeCounts), > (68056473384187692692674921486353642290,102084710076281539039012382229530463435]> > > > > Now, interestingly that last-line indicates that the endpoint for the > repair is the node I'm running the repair on. I restarted and tried > the repair again, and it immediately tried the endpoint of itself, and > did not progress any further. > > > Should a node ever make an AEService tree request with itself as the endpoint? > > > Thanks, > Jason >
