On Wed, 2021-01-20 at 08:56 +0800, Xiubo Li wrote: > On 2021/1/18 19:08, Jeff Layton wrote: > > On Mon, 2021-01-18 at 17:10 +0800, Xiubo Li wrote: > > > On 2021/1/13 5:48, Jeff Layton wrote: > > > > On Sun, 2021-01-10 at 10:01 +0800, xiubli@xxxxxxxxxx wrote: > > > > > From: Xiubo Li <xiubli@xxxxxxxxxx> > > > > > > > > > > If the Fb cap is used it means the current inode is flushing the > > > > > dirty data to OSD, just defer flushing the capsnap. > > > > > > > > > > URL: https://tracker.ceph.com/issues/48679 > > > > > URL: https://tracker.ceph.com/issues/48640 > > > > > Signed-off-by: Xiubo Li <xiubli@xxxxxxxxxx> > > > > > --- > > > > > > > > > > V3: > > > > > - Add more comments about putting the inode ref > > > > > - A small change about the code style > > > > > > > > > > V2: > > > > > - Fix inode reference leak bug > > > > > > > > > > Â Â fs/ceph/caps.c | 32 +++++++++++++++++++------------- > > > > > Â Â fs/ceph/snap.c | 6 +++--- > > > > > Â Â 2 files changed, 22 insertions(+), 16 deletions(-) > > > > > > > > > Hi Xiubo, > > > > > > > > This patch seems to cause hangs in some xfstests (generic/013, in > > > > particular). I'll take a closer look when I have a chance, but I'm > > > > dropping this for now. > > > Okay. > > > > > > BTW, what's your test commands to reproduce it ? I will take a look when > > > I am free these days or later. > > > > > > BRs > > > > > I set up xfstests to run on cephfs, and then just run: > > > > $ sudo ./check generic/013 > > > > It wouldn't reliably complete with this patch in place. Setting up > > xfstests is the "hard part". I'll plan to roll up a wiki page on how to > > do that soon (that's good info to have out there anyway). > > Okay, sure. > I'm not sure where this should be documented. Still, here's a local.config that I'm using now (with comments). I'm happy to merge this somewhere for posterity, but not sure where it should go. -- Jeff Layton <jlayton@xxxxxxxxxx>
# # For running xfstests on kcephfs # # In this example, we've created two different named filesystems: "test" # and "scratch. They must be pre-created on the ceph cluster before the # test is run. # # Standard mountpoint locations are fine # export TEST_DIR=/mnt/test export SCRATCH_MNT=/mnt/scratch # # "check" can't automatically detect ceph device strings, so we must # explicitly declare that we want to use "-t ceph". # export FSTYP=ceph # # The check script gets very confused when two different mounts use # the same device string. Eventually we may fix this in ceph so we can # get monaddrs from the config, but for now, we must declare the location # of the mons explicitly. Note that we're using two different monaddrs # here, though these are using the same cluster. The monaddrs must also # match the type of ms_mode option that is in effect (i.e. # ms_mode=legacy requires v1 monaddrs). # export TEST_DEV=10.10.10.1:3300:/ export SCRATCH_DEV=10.10.10.2:3300:/ # # TEST_FS_MOUNT_OPTS is for /mnt/test, and MOUNT_OPTONS is for /mnt/scratch # # Here, we're using the admin account for both mounts. The credentials # should be in a standard keyring location somewhere. See: # # https://docs.ceph.com/en/latest/rados/operations/user-management/#keyring-management # COMMON_OPTIONS="name=admin,ms_mode=crc" # # asyncronous directory ops # COMMON_OPTIONS+=",nowsync" # # swizzle in the COMMON_OPTIONS # TEST_FS_MOUNT_OPTS="-o ${COMMON_OPTIONS},mds_namespace=test" MOUNT_OPTIONS="-o ${COMMON_OPTIONS},mds_namespace=scratch" # # fscache -- this needs a different option for each # # TEST_FS_MOUNT_OPTS+=",fsc=test" # MOUNT_OPTIONS+=",fsc=scratch" export TEST_FS_MOUNT_OPTS export MOUNT_OPTIONS
