On Fri, Oct 5, 2012 at 2:30 PM, Simon Willnauer <[email protected]> wrote: > On Fri, Oct 5, 2012 at 8:27 PM, Michael McCandless > <[email protected]> wrote: >> I committed a fix ... this was actually a real bug, not a test bug, >> and my commit is just a workaround. >> >> There seems to be a starvation issue that can allow multiple indexing >> threads to flush segments faster than the single thread can "publish" >> them. I think the problem is that "publishing" is actually somewhat >> costly (creates CFS, writes .si file, etc.), so somehow we need to >> make these steps concurrent too ... I'll open an issue for a longer >> term fix. > hoho! looking forward to that!
I'm just going to open the issue :) Not sure how to fix it ... Mike McCandless http://blog.mikemccandless.com --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
