On Sat, 2013-10-05 at 11:39 +0200, Sebastiano Vigna wrote: > On 25 Sep 2013, at 2:00 PM, Oleg Kalnichevski <[email protected]> wrote: > > > I do not have any suggestions other than taking note of problematic > > hosts, monitoring them closely and probably applying more aggressive > > timeout parameters to those hosts by dynamically reducing socket timeout > > if throughput drops below a certain limit. > > > We did it, and it solved part of the problems we're having. > > The fact is that there are parts of the stack we cannot control. This threads > are stuck since ~12h (we have about 30 of the same kinds over 6000 overall > fetching threads on 3 machines):
... > Suggestions? Should we give an interrupt after a certain amount of time has > elapsed? > Well, that can certainly be seen as the last resort if everything else fails. Oleg --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
