Skip to main content

You are not logged in. Your edit will be placed in a queue until it is peer reviewed.

We welcome edits that make the post easier to understand and more valuable for readers. Because community members review edits, please try to make the post substantially better than how you found it, for example, by fixing grammar or adding additional resources and hyperlinks.

3
  • OK... I don't know. I think that eliminating the journal has changed the setup. I updated the tests after that as I wanted to have more than one output from IOStat. But the problem remains. At the end of a crawl we move some data using the transfer() facility of Java NIO. This means that we move from file to file 30G with just one method call. It's going on since days (literally) instead of a few minutes, copying a few hundreds K per second. The weird thing is that while that copy is going on so low, a dd on the same disk now does 130MB/s.
    – seba
    Commented Apr 22, 2014 at 8:37
  • Another thing that happened is that also a second server started to behave similarly, and we discovered that the two bad server have an old (2.8.0) BIOS w.r.t. the new ones (3.3.2). So we're going to update and retry.
    – seba
    Commented Apr 22, 2014 at 8:38
  • And... no, it's not the BIOS. At this point I'm thinking of some very low-level hardware problem, like interrupts timing out or something. Our next try will be reformatting the large partitions and test again.
    – seba
    Commented Apr 22, 2014 at 10:45