Date: Fri, 15 Jun 2012 11:34:48 -0400 (EDT) From: Rick Macklem <rmacklem@uoguelph.ca> To: Konstantin Belousov <kostikbel@gmail.com> Cc: freebsd-fs@freebsd.org, Pavlo <devgs@ukr.net> Subject: Re: mmap() incoherency on hi I/O load (FS is zfs) Message-ID: <1116727909.1836239.1339774488001.JavaMail.root@erie.cs.uoguelph.ca> In-Reply-To: <20120614122456.GZ2337@deviant.kiev.zoral.com.ua>
next in thread | previous in thread | raw e-mail | index | archive | help
Kostik wrote: > On Thu, Jun 14, 2012 at 07:32:36AM -0400, Rick Macklem wrote: > > Pavlo wrote: > > > There's a case when some parts of files that are mapped and then > > > modified getting corrupted. By corrupted I mean some data is ok > > > (one > > > that > > > was written using write()/pwrite()) but some looks like it never > > > existed. > > > Like it was some time in buffers, when several processes > > > simultaneously > > > (of course access was synchronised) used shared pages and reported > > > it's > > > existence. But after time pass they (processes) screamed that it > > > is > > > now > > > lost. Only part of data written with pwrite() was there. > > > Everything > > > that > > > was written via mmap() is zero. > > > > > > So as I said it occurs on hi I/O busyness. When in background 4+ > > > processes do indexing of huge ammount of data. Also I want to > > > note, it > > > never occurred in the life of our project while we used mmap() > > > under > > > same I/O stress conditions when mapping was done for a whole file > > > of > > > just > > > a part(header) starting from a beginning of a file. First time we > > > used > > > mapping of individual pages, just to save RAM, and this popped up. > > > > > > Solution for this problem is msync() before any munmap(). But man > > > says: > > > > > > The msync() system call is usually not needed since BSD implements > > > a > > > coherent file system buffer cache. However, it may be used to > > > associate > > > dirty VM pages with file system buffers and thus cause them to be > > > flushed > > > to physical media sooner rather than later. > > > > > > Any thoughts? Thanks. > > > > > With a recent kernel from head, I am seeing dirty mmap'd pages being > > written > > quite late for the NFSv4 client. Even after the NFS client > > VOP_RECLAIM() has > > been called, it seems. I didn't observe this behaviour in a kernel > > from > > head in March. (I don't know enough about the vm/mmap area to know > > if this > > is correct behaviour or not?) > > > > I thought I'd mention this, since you didn't say how recent a kernel > > you > > were running and thought it might be caused by the same change? > Can you, please, comment more on this ? > How is this possible at all ? > > Could you please show at least a backtrace for the moment when a write > request is made for the page which belong to already reclaimed vnode ? After some off list discussion, it was determined that my problem was doing nfsrpc_close() before vnode_destroy_object() in the NFSv4 client's VOP_RECLAIM(). This is an NFSv4 specific bug and wouldn't be related to the above issue. Sorry about the noise, rick
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1116727909.1836239.1339774488001.JavaMail.root>