ztest spa_vdev_add() returns EBUSY

[behlendorf]

Very, very, very rarely do I observe the following running ztest.

5 vdevs, 7 datasets, 23 threads, 300 seconds...
Pass   1,  SIGKILL,   0 ENOSPC,  8.2% of  238M used,  17% done,    4m08s to go
Pass   2,  SIGKILL,   0 ENOSPC, 15.9% of  238M used,  35% done,    3m13s to go
Pass   3,  SIGKILL,   0 ENOSPC, 16.9% of  238M used,  41% done,    2m57s to go
Pass   4,  SIGKILL,   0 ENOSPC, 20.6% of  238M used,  59% done,    2m03s to go
Pass   5,  SIGKILL,   0 ENOSPC, 20.4% of  238M used,  70% done,    1m30s to go
ztest: spa_vdev_add(0x7fccdc0f98e0) = 16
child died with signal 6
program finished with exit code 3
elapsedTime=624.433468

[birdietwin]

A build on Centos 6.2 kernel 2.6.32-220.4.2.el6.x86_64 (AWS EC2 ami-5bf15d32) has an error in step 10 test 4.

[root@centos-ami ec2-user]# /usr/libexec/zfs/zconfig.sh -c
1 persistent zpool.cache Pass
2 scan disks for pools to import Pass
3 zpool import/export device Pass
4 zpool insmod/rmmod device Pass
5 zvol+ext2 volume done Pass
6 zvol+ext2 snapshot done Pass
7 zvol+ext2 clone done Pass
8 zfs send/receive done Pass
9 zpool events Pass
10 zpool add/remove vdev Fail (4)

Looking at the zconfig.sh test 4 is:

...

Add and remove a cache vdev by full path

    zconfig_add_vdev ${POOL_NAME} cache ${SDDEVICE} || fail 4

...

Same issue?

[behlendorf]

This issue also is believed to be fixed. We can reopen it if needed by I haven't seen this occur in years.