Re: Packet loss every 30.999 seconds
- From: David G Lawrence <dg@xxxxxxxxxxxxxx>
- Date: Mon, 17 Dec 2007 02:39:36 -0800
One more comment on my last email... The patch that I included is not
meant as a real fix - it is just a bandaid. The real problem appears to
be that a very large number of vnodes (all of them?) are getting synced
(i.e. calling ffs_syncvnode()) every time. This should normally only
happen for dirty vnodes. I suspect that something is broken with this
check:
if (vp->v_type == VNON || ((ip->i_flag &
(IN_ACCESS | IN_CHANGE | IN_MODIFIED | IN_UPDATE)) == 0 &&
vp->v_bufobj.bo_dirty.bv_cnt == 0)) {
VI_UNLOCK(vp);
continue;
}
...like the i_flag flags aren't ever getting properly cleared (or bv_cnt
is always non-zero).
...but I don't have the time to chase this down.
-DG
David G. Lawrence
President
Download Technologies, Inc. - http://www.downloadtech.com - (866) 399 8500
The FreeBSD Project - http://www.freebsd.org
Pave the road of life with opportunities.
_______________________________________________
freebsd-net@xxxxxxxxxxx mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-net
To unsubscribe, send any mail to "freebsd-net-unsubscribe@xxxxxxxxxxx"
- Follow-Ups:
- Re: Packet loss every 30.999 seconds
- From: Bruce Evans
- Re: Packet loss every 30.999 seconds
- References:
- Packet loss every 30.999 seconds
- From: Mark Fullmer
- Packet loss every 30.999 seconds
- Prev by Date: Re: Added native socks support to libc in FreeBSD 7
- Next by Date: Re: Packet loss every 30.999 seconds
- Previous by thread: Packet loss every 30.999 seconds
- Next by thread: Re: Packet loss every 30.999 seconds
- Index(es):
Relevant Pages
- Re: Packet loss every 30.999 seconds
... The patch that I included is not ... meant as a real fix - it is just
a bandaid. ... happen for dirty vnodes. ... (freebsd-stable) - Re: pty_chars_in_buffer NULL pointer (kernel oops)
... This is the hacky "hide the problem" patch that is in my current tree (and ...
in theory some line discipline might do something strange in ... # Fix possible pty line
discipline race. ... Real fix under discussion. ... (Linux-Kernel) - Re: [RFD] Explicitly documenting patch submission
... patch, delete the diff when I'm done etc. ... as a basis for the real fix.
... As asking submitters to sign off on modified versions ... send the line "unsubscribe
linux-kernel" in ... (Linux-Kernel) - Re: kernel BUG at mm/rmap.c:480 in 2.6.10-rc3-bk7
... On Thu, 16 Dec 2004, Andrea Arcangeli wrote: ... I still don't excude they're
a real fix though, ... a stupid stupid bug in my patch: I forget to delete the PageReserved
... static void zap_pte_range(struct mmu_gather *tlb, ... (Linux-Kernel) - Re: [patch 2.6.12-rc2] revert fs/char_dev.c CONFIG_BASE_FULL change
... > a patch from Matt Mackall that misbehaves. ... I actually asked Linus to
revert it right ... I've got a Real Fix (a refactor of block and chardev name ...
send the line "unsubscribe linux-kernel" in ... (Linux-Kernel)