There are some issues with the way counter values are collected and summarized that cause this display behavior. It partially has to do with speculative execution. I believe it is fixed in the later versions of 0.19 and beyond. I can't remember or find the jira associated with this.
On Sun, Dec 27, 2009 at 2:28 AM, Jeff Zhang <zjf...@gmail.com> wrote: > No, if you look at the click the job id and watch the details of this job, > some task must failed. > > > Jeff Zhang > > 2009/12/27 Doopah Shaf <doopha.s...@gmail.com> > > > The web UI doesn't show any failures. > > > > And it does show several reduce tasks with progress = 475.70%... any > clues? > > > > On Sun, Dec 27, 2009 at 12:20 PM, <lamfeeli...@gmail.com> wrote: > > > > > Task failed? > > > > > > ---- 原始信息 ---- > > > 发自: Doopah Shaf <doopha.s...@gmail.com> > > > 发送时间: > > > 收件人: common-user@hadoop.apache.org > > > 主题: hadoop job progress going back > > > > > > Running jobs sometimes displays progress as follows: > > > > > > map 100% reduce 10% > > > > > > and then > > > > > > map 100% reduce 0% > > > > > > What causes this? Is this speculative execution restarting tasks? > > > > > > Cheers, > > > DS > > > > > > -- Pro Hadoop, a book to guide you from beginner to hadoop mastery, http://www.amazon.com/dp/1430219424?tag=jewlerymall www.prohadoopbook.com a community for Hadoop Professionals
