Peter Taphouse wrote: > Hi, > > I'm currently testing an iscsi box based on the gigastorATX board. When > running a couple of basic bonnie benchmark over the target the write > portion of the test runs fine up until the re-write starts, at which > point everything slows down and no i/o happens for periods of up to a > minute followed by burts of reads and writes. > > Taking this back to the most basic thing I could think of I've decided > to just run cat /dev/sdXXX > /dev/zero . This can pause for well over a > minute before any i/o appears to happen. > > I'm running the version of 2.0-865 of the initiator, on a 64bit debian > etch install, over intel gige with an 8000 byte mtu. I've tried > identical tests with an open-e based target and have encountered no > problems. I've also tried the standard version of the initiator that > comes packaged with etch, and that behaved the same as the latest version. > > Has anyone seen anything similar before, or can shed any light on this?
Are you using exactly 2.0-865? If so could you try the userspace tools and kernel modules in: http://www.open-iscsi.org/bits/open-iscsi-2.0-865.15.tar.gz There are two write fixes in there. It sounds like you are hitting a problem where we counted some variable incorrectly and halt writes for no good reason. --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "open-iscsi" group. To post to this group, send email to firstname.lastname@example.org To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/open-iscsi -~----------~----~----~----~------~----~------~--~---