vdev probe to slow disk can stall mmp write checker

Simplify vdev probes in the zio_vdev_io_done context to
avoid holding the spa config lock for a long duration.

Also allow zpool clear if no evidence of another host
is using the pool.

Sponsored-by: Klara, Inc.
Sponsored-by: Wasabi Technology, Inc.
Reviewed-by: Olaf Faaland <faaland1@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Don Brady <don.brady@klarasystems.com>
Closes #15839
This commit is contained in:
Don Brady
2024-04-29 15:35:53 -06:00
committed by GitHub
parent b28461b7c6
commit c3f2f1aa2d
16 changed files with 242 additions and 52 deletions
+3 -2
View File
@@ -664,12 +664,13 @@ mmp_thread(void *arg)
(gethrtime() - mmp->mmp_last_write) > mmp_fail_ns) {
zfs_dbgmsg("MMP suspending pool '%s': gethrtime %llu "
"mmp_last_write %llu mmp_interval %llu "
"mmp_fail_intervals %llu mmp_fail_ns %llu",
"mmp_fail_intervals %llu mmp_fail_ns %llu txg %llu",
spa_name(spa), (u_longlong_t)gethrtime(),
(u_longlong_t)mmp->mmp_last_write,
(u_longlong_t)mmp_interval,
(u_longlong_t)mmp_fail_intervals,
(u_longlong_t)mmp_fail_ns);
(u_longlong_t)mmp_fail_ns,
(u_longlong_t)spa->spa_uberblock.ub_txg);
cmn_err(CE_WARN, "MMP writes to pool '%s' have not "
"succeeded in over %llu ms; suspending pool. "
"Hrtime %llu",