Project

General

Profile

Actions

Bug #5367

closed

multiclient tests: kernel mount gets EPERM

Added by Sage Weil almost 11 years ago. Updated over 10 years ago.

Status:
Resolved
Priority:
High
Assignee:
-
Category:
-
Target version:
-
% Done:

0%

Source:
Q/A
Tags:
Backport:
Regression:
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Component(FS):
Labels (FS):
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

2013-06-15T11:53:25.609 INFO:teuthology.orchestra.run.out:access    bw(MiB/s)  block(KiB) xfer(KiB)  open(s)    wr/rd(s)   close(s)   total(s)   iter
2013-06-15T11:53:25.609 INFO:teuthology.orchestra.run.out:------    ---------  ---------- ---------  --------   --------   --------   --------   ----
2013-06-15T11:53:25.611 INFO:teuthology.orchestra.run.out:ior ERROR: open64() failed, errno 13, Permission denied (aiori-POSIX.c:158)
2013-06-15T11:53:25.611 INFO:teuthology.orchestra.run.out:ior ERROR: open64() failed, errno 13, Permission denied (aiori-POSIX.c:158)
2013-06-15T11:53:25.611 INFO:teuthology.orchestra.run.out:ior ERROR: open64() failed, errno 13, Permission denied (aiori-POSIX.c:158)
ubuntu@teuthology:/a/sage-2013-06-15_09:42:09-foo-next-testing-basic/36504$ cat orig.config.yaml 
kernel:
  kdb: true
  sha1: f28d6bf9ef4d3f8dd2e669fd21be827663f215e2
machine_type: plana
nuke-on-error: true
overrides:
  ceph:
    conf:
      mon:
        debug mon: 20
        debug ms: 20
        debug paxos: 20
      osd:
        osd op thread timeout: 60
    fs: btrfs
    log-whitelist:
    - slow request
    sha1: 9253cd8174ca7690414923259b01c0a484323f19
  install:
    ceph:
      sha1: 9253cd8174ca7690414923259b01c0a484323f19
  s3tests:
    branch: next
  workunit:
    sha1: 9253cd8174ca7690414923259b01c0a484323f19
roles:
- - mon.a
  - mon.b
  - mon.c
  - mds.a
  - osd.0
  - osd.1
- - client.2
- - client.1
- - client.0
tasks:
- chef: null
- clock.check: null
- install: null
- ceph: null
- kclient: null
- pexec:
    clients:
    - cd $TESTDIR
    - wget http://ceph.com/qa/ior.tbz2
    - tar xvfj ior.tbz2
    - cd ior
    - ./configure
    - make
    - make install DESTDIR=$TESTDIR/binary/
    - cd $TESTDIR/
    - rm ior.tbz2
    - rm -r ior
    - ln -s $TESTDIR/mnt.* $TESTDIR/gmnt
- ssh_keys: null
- mpi:
    exec: $TESTDIR/binary/usr/local/bin/ior -e -w -r -W -b 10m -a POSIX -o $TESTDIR/gmnt/ior.testfile
- pexec:
    all:
    - rm -f $TESTDIR/gmnt/ior.testfile
    - rm -f $TESTDIR/gmnt


Related issues 1 (0 open1 closed)

Related to CephFS - Bug #5761: teuthology: MPI test sometimes fails with a permission deniedResolved07/25/2013

Actions
Actions #1

Updated by Sage Weil almost 11 years ago

mpi-fsx also gets EPERM.

Actions #2

Updated by Zheng Yan almost 11 years ago

kclient and MDS never return -EACCES. was ior executed with root privilege?

Actions #3

Updated by Sage Weil almost 11 years ago

saw similar error on ceph-fuse:

2013-07-17T12:35:23.001 INFO:teuthology.orchestra.run.out:access    bw(MiB/s)  block(KiB) xfer(KiB)  open(s)    wr/rd(s)   close(s)   total(s)   iter
2013-07-17T12:35:23.001 INFO:teuthology.orchestra.run.out:------    ---------  ---------- ---------  --------   --------   --------   --------   ----
2013-07-17T12:35:23.340 INFO:teuthology.orchestra.run.out:ior ERROR: open64() failed, errno 13, Permission denied (aiori-POSIX.c:158)
2013-07-17T12:35:23.340 INFO:teuthology.orchestra.run.err:application called MPI_Abort(MPI_COMM_WORLD, -1) - process 2
2013-07-17T12:35:23.843 INFO:teuthology.orchestra.run.err:Fatal error in PMPI_Reduce: Other MPI error, error stack:
2013-07-17T12:35:23.843 INFO:teuthology.orchestra.run.err:PMPI_Reduce(1270)...............: MPI_Reduce(sbuf=0x7fff84fe1ffc, rbuf=0x7fff84fe2008, count=1, MPI_INT, MPI_SUM, root=0, comm=0x84000000) failed
2013-07-17T12:35:23.843 INFO:teuthology.orchestra.run.err:MPIR_Reduce_impl(1087)..........:
2013-07-17T12:35:23.843 INFO:teuthology.orchestra.run.err:MPIR_Reduce_intra(895)..........:
2013-07-17T12:35:23.843 INFO:teuthology.orchestra.run.err:MPIR_Reduce_binomial(144).......:
2013-07-17T12:35:23.843 INFO:teuthology.orchestra.run.err:MPIDI_CH3U_Recvq_FDU_or_AEP(380): Communication error with rank 2

job was
ubuntu@teuthology:/a/teuthology-2013-07-17_01:01:03-fs-next-testing-basic/70462$ cat orig.config.yaml 
kernel:
  kdb: true
  sha1: 77c8bf2f972a9d6ff446c49a41678bf931bbee44
machine_type: plana
nuke-on-error: true
overrides:
  admin_socket:
    branch: next
  ceph:
    conf:
      mon:
        debug mon: 20
        debug ms: 20
        debug paxos: 20
      osd:
        osd op thread timeout: 60
    fs: btrfs
    log-whitelist:
    - slow request
    sha1: 884fa2fcb6d707b23317bab1da909586ddc27608
  ceph-deploy:
    conf:
      client:
        debug monc: 20
        debug ms: 1
        debug objecter: 20
        debug rados: 20
        log file: /var/log/ceph/ceph-..log
      mon:
        debug mon: 20
        debug ms: 20
        debug paxos: 20
  install:
    ceph:
      sha1: 884fa2fcb6d707b23317bab1da909586ddc27608
  s3tests:
    branch: next
  workunit:
    sha1: 884fa2fcb6d707b23317bab1da909586ddc27608
roles:
- - mon.a
  - mon.b
  - mon.c
  - mds.a
  - osd.0
  - osd.1
  - osd.2
- - client.2
- - client.1
- - client.0
tasks:
- chef: null
- clock.check: null
- install: null
- ceph: null
- ceph-fuse: null
- pexec:
    clients:
    - cd $TESTDIR
    - wget http://ceph.com/qa/ior.tbz2
    - tar xvfj ior.tbz2
    - cd ior
    - ./configure
    - make
    - make install DESTDIR=$TESTDIR/binary/
    - cd $TESTDIR/
    - rm ior.tbz2
    - rm -r ior
    - ln -s $TESTDIR/mnt.* $TESTDIR/gmnt
- ssh_keys: null
- mpi:
    exec: $TESTDIR/binary/usr/local/bin/ior -e -w -r -W -b 10m -a POSIX -o $TESTDIR/gmnt/ior.testfile
- pexec:
    all:
    - rm -f $TESTDIR/gmnt/ior.testfile
    - rm -f $TESTDIR/gmnt
    - rm -rf $TESTDIR/binary

Actions #4

Updated by Sage Weil almost 11 years ago

Zheng Yan wrote:

kclient and MDS never return -EACCES. was ior executed with root privilege?

not root...

Actions #5

Updated by Zheng Yan over 10 years ago

can't reproduce locally. Is it possible your clients have different 'username -> uid' mapping?

Actions #6

Updated by Sage Weil over 10 years ago

ah, a few do! the ones this test failed on don't, but maybe they were reimaged since then.

Actions #7

Updated by Greg Farnum over 10 years ago

  • Status changed from New to Resolved

We can reopen if we see this again, but the tests are now checking (and chef is setting) the uid for the ubuntu (test) user.

Actions

Also available in: Atom PDF