evin/ms-nfs41-client
3
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

recovery: recover from STALE_STATEID errors

Browse source 
consider an operation that takes a stateid, and results in a BADSESSION error due to server reboot.  we'll recover the client and session, and send OPENs to reclaim all of the client's state.  but after recovery, we'll resend the original operation with the original stateid, and this will result in a STALE_STATEID error

we handle this by making use of the information in stateid_arg.  if we determine that stateid_arg.stateid is different from the nfs41_open_state's stateid, we copy the new stateid into stateid_arg.stateid and retry

note that if another thread is in recovery, it hasn't finished reclaiming its open state yet; so we wait on recovery to finish before comparing the stateids

Signed-off-by: Casey Bodley <cbodley@citi.umich.edu>
This commit is contained in:
Casey Bodley 2010-11-15 15:24:14 -05:00 • committed by unknown
parent 3ecd38e414
commit 8616b03597
1 changed files with 73 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

73
daemon/nfs41_compound.c
View file

@ -253,6 +253,79 @@ retry:
nfs41_renew_in_progress(session->client, &zero); nfs41_renew_in_progress(session->client, &zero);
goto do_retry; goto do_retry;
case NFS4ERR_STALE_STATEID:
if (compound->args.argarray[0].op == OP_SEQUENCE) {
nfs41_sequence_args *seq = (nfs41_sequence_args*)
compound->args.argarray[0].arg;
nfs41_session_free_slot(session, seq->sa_slotid);
}
{
nfs_argop4 *argop = &compound->args.argarray[
compound->res.resarray_count-1];
stateid_arg *stateid = NULL;
if (argop->op == OP_CLOSE) {
nfs41_op_close_args *close = (nfs41_op_close_args*)argop->arg;
stateid = close->stateid;
} else if (argop->op == OP_READ) {
nfs41_read_args *read = (nfs41_read_args*)argop->arg;
stateid = read->stateid;
} else if (argop->op == OP_WRITE) {
nfs41_write_args *write = (nfs41_write_args*)argop->arg;
stateid = write->stateid;
} else if (argop->op == OP_LOCK) {
nfs41_lock_args *lock = (nfs41_lock_args*)argop->arg;
if (lock->locker.new_lock_owner)
stateid = lock->locker.u.open_owner.open_stateid;
else
stateid = lock->locker.u.lock_owner.lock_stateid;
} else if (argop->op == OP_LOCKU) {
nfs41_locku_args *locku = (nfs41_locku_args*)argop->arg;
stateid = locku->lock_stateid;
} else if (argop->op == OP_SETATTR) {
nfs41_setattr_args *setattr = (nfs41_setattr_args*)argop->arg;
stateid = setattr->stateid;
} else if (argop->op == OP_LAYOUTGET) {
pnfs_layoutget_args *lget = (pnfs_layoutget_args*)argop->arg;
stateid = lget->stateid;
}
if (stateid) {
bool_t retry = FALSE;
switch (stateid->type) {
case STATEID_OPEN:
/* if there's recovery in progress, wait for it to finish */
while (nfs41_renew_in_progress(session->client, NULL)) {
DWORD wait = WaitForSingleObject(session->client->cond, INFINITE);
if (wait != WAIT_OBJECT_0) {
print_condwait_status(1, wait);
break;
}
}
/* if the open stateid is different, update and retry */
AcquireSRWLockShared(&stateid->open->lock);
if (memcmp(&stateid->stateid, &stateid->open->stateid, sizeof(stateid4))) {
memcpy(&stateid->stateid, &stateid->open->stateid, sizeof(stateid4));
retry = TRUE;
}
ReleaseSRWLockShared(&stateid->open->lock);
break;
default:
eprintf("%s returned %s: can't recover stateid type %u\n",
nfs_opnum_to_string(argop->op),
nfs_error_string(compound->res.status), stateid->type);
break;
}
if (retry) {
dprintf(1, "Stateid has been reclaimed, retrying..\n");
goto do_retry;
}
}
}
goto out;
case NFS4ERR_GRACE: case NFS4ERR_GRACE:
case NFS4ERR_DELAY: case NFS4ERR_DELAY:
#define RETRY_INDEFINITELY #define RETRY_INDEFINITELY