* Copyright (c) 2003, 2010, Oracle and/or its affiliates. All rights reserved.
* Use is subject to license terms.
*
- * Copyright (c) 2011, 2014, Intel Corporation.
+ * Copyright (c) 2011, 2015, Intel Corporation.
*/
/*
* This file is part of Lustre, http://www.lustre.org/
"still on delayed list");
}
- CERROR("%s: RPCs in \"%s\" phase found (%d). "
+ CERROR("%s: Unregistering RPCs found (%d). "
"Network is sluggish? Waiting them "
"to error out.\n", cli_tgt,
- ptlrpc_phase2str(RQ_PHASE_UNREGISTERING),
atomic_read(&imp->imp_unregistering));
}
spin_unlock(&imp->imp_lock);
}
list_for_each_entry(conn, &imp->imp_conn_list, oic_item) {
- CDEBUG(D_HA, "%s: connect to NID %s last attempt "LPU64"\n",
+ CDEBUG(D_HA, "%s: connect to NID %s last attempt %llu\n",
imp->imp_obd->obd_name,
libcfs_nid2str(conn->oic_conn->c_peer.nid),
conn->oic_last_attempt);
*/
int ptlrpc_connect_import(struct obd_import *imp)
{
- struct obd_device *obd = imp->imp_obd;
- int initial_connect = 0;
- int set_transno = 0;
- __u64 committed_before_reconnect = 0;
- struct ptlrpc_request *request;
- char *bufs[] = { NULL,
- obd2cli_tgt(imp->imp_obd),
- obd->obd_uuid.uuid,
- (char *)&imp->imp_dlm_handle,
- (char *)&imp->imp_connect_data };
- struct ptlrpc_connect_async_args *aa;
- int rc;
- ENTRY;
+ struct obd_device *obd = imp->imp_obd;
+ int initial_connect = 0;
+ int set_transno = 0;
+ __u64 committed_before_reconnect = 0;
+ struct ptlrpc_request *request;
+ char *bufs[] = { NULL,
+ obd2cli_tgt(imp->imp_obd),
+ obd->obd_uuid.uuid,
+ (char *)&imp->imp_dlm_handle,
+ (char *)&imp->imp_connect_data };
+ struct ptlrpc_connect_async_args *aa;
+ int rc;
+ ENTRY;
spin_lock(&imp->imp_lock);
if (imp->imp_state == LUSTRE_IMP_CLOSED) {
spin_unlock(&imp->imp_lock);
CERROR("already connected\n");
RETURN(0);
- } else if (imp->imp_state == LUSTRE_IMP_CONNECTING) {
+ } else if (imp->imp_state == LUSTRE_IMP_CONNECTING ||
+ imp->imp_connected) {
spin_unlock(&imp->imp_lock);
CERROR("already connecting\n");
RETURN(-EALREADY);
}
- IMPORT_SET_STATE_NOLOCK(imp, LUSTRE_IMP_CONNECTING);
+ IMPORT_SET_STATE_NOLOCK(imp, LUSTRE_IMP_CONNECTING);
- imp->imp_conn_cnt++;
- imp->imp_resend_replay = 0;
+ imp->imp_conn_cnt++;
+ imp->imp_resend_replay = 0;
- if (!lustre_handle_is_used(&imp->imp_remote_handle))
- initial_connect = 1;
- else
- committed_before_reconnect = imp->imp_peer_committed_transno;
+ if (!lustre_handle_is_used(&imp->imp_remote_handle))
+ initial_connect = 1;
+ else
+ committed_before_reconnect = imp->imp_peer_committed_transno;
- set_transno = ptlrpc_first_transno(imp,
- &imp->imp_connect_data.ocd_transno);
+ set_transno = ptlrpc_first_transno(imp,
+ &imp->imp_connect_data.ocd_transno);
spin_unlock(&imp->imp_lock);
- rc = import_select_connection(imp);
- if (rc)
- GOTO(out, rc);
+ rc = import_select_connection(imp);
+ if (rc)
+ GOTO(out, rc);
rc = sptlrpc_import_sec_adapt(imp, NULL, NULL);
- if (rc)
- GOTO(out, rc);
+ if (rc)
+ GOTO(out, rc);
- /* Reset connect flags to the originally requested flags, in case
- * the server is updated on-the-fly we will get the new features. */
- imp->imp_connect_data.ocd_connect_flags = imp->imp_connect_flags_orig;
+ /* Reset connect flags to the originally requested flags, in case
+ * the server is updated on-the-fly we will get the new features. */
+ imp->imp_connect_data.ocd_connect_flags = imp->imp_connect_flags_orig;
+ imp->imp_connect_data.ocd_connect_flags2 = imp->imp_connect_flags2_orig;
/* Reset ocd_version each time so the server knows the exact versions */
imp->imp_connect_data.ocd_version = LUSTRE_VERSION_CODE;
- imp->imp_msghdr_flags &= ~MSGHDR_AT_SUPPORT;
- imp->imp_msghdr_flags &= ~MSGHDR_CKSUM_INCOMPAT18;
-
- rc = obd_reconnect(NULL, imp->imp_obd->obd_self_export, obd,
- &obd->obd_uuid, &imp->imp_connect_data, NULL);
- if (rc)
- GOTO(out, rc);
-
- request = ptlrpc_request_alloc(imp, &RQF_MDS_CONNECT);
- if (request == NULL)
- GOTO(out, rc = -ENOMEM);
-
- rc = ptlrpc_request_bufs_pack(request, LUSTRE_OBD_VERSION,
- imp->imp_connect_op, bufs, NULL);
- if (rc) {
- ptlrpc_request_free(request);
- GOTO(out, rc);
- }
+ imp->imp_msghdr_flags &= ~MSGHDR_AT_SUPPORT;
+ imp->imp_msghdr_flags &= ~MSGHDR_CKSUM_INCOMPAT18;
+
+ rc = obd_reconnect(NULL, imp->imp_obd->obd_self_export, obd,
+ &obd->obd_uuid, &imp->imp_connect_data, NULL);
+ if (rc)
+ GOTO(out, rc);
- /* Report the rpc service time to the server so that it knows how long
- * to wait for clients to join recovery */
- lustre_msg_set_service_time(request->rq_reqmsg,
- at_timeout2est(request->rq_timeout));
-
- /* The amount of time we give the server to process the connect req.
- * import_select_connection will increase the net latency on
- * repeated reconnect attempts to cover slow networks.
- * We override/ignore the server rpc completion estimate here,
- * which may be large if this is a reconnect attempt */
- request->rq_timeout = INITIAL_CONNECT_TIMEOUT;
- lustre_msg_set_timeout(request->rq_reqmsg, request->rq_timeout);
-
- request->rq_no_resend = request->rq_no_delay = 1;
- request->rq_send_state = LUSTRE_IMP_CONNECTING;
- /* Allow a slightly larger reply for future growth compatibility */
- req_capsule_set_size(&request->rq_pill, &RMF_CONNECT_DATA, RCL_SERVER,
- sizeof(struct obd_connect_data)+16*sizeof(__u64));
- ptlrpc_request_set_replen(request);
- request->rq_interpret_reply = ptlrpc_connect_interpret;
-
- CLASSERT(sizeof (*aa) <= sizeof (request->rq_async_args));
- aa = ptlrpc_req_async_args(request);
- memset(aa, 0, sizeof *aa);
-
- aa->pcaa_peer_committed = committed_before_reconnect;
- aa->pcaa_initial_connect = initial_connect;
-
- if (aa->pcaa_initial_connect) {
+ request = ptlrpc_request_alloc(imp, &RQF_MDS_CONNECT);
+ if (request == NULL)
+ GOTO(out, rc = -ENOMEM);
+
+ rc = ptlrpc_request_bufs_pack(request, LUSTRE_OBD_VERSION,
+ imp->imp_connect_op, bufs, NULL);
+ if (rc) {
+ ptlrpc_request_free(request);
+ GOTO(out, rc);
+ }
+
+ /* Report the rpc service time to the server so that it knows how long
+ * to wait for clients to join recovery */
+ lustre_msg_set_service_time(request->rq_reqmsg,
+ at_timeout2est(request->rq_timeout));
+
+ /* The amount of time we give the server to process the connect req.
+ * import_select_connection will increase the net latency on
+ * repeated reconnect attempts to cover slow networks.
+ * We override/ignore the server rpc completion estimate here,
+ * which may be large if this is a reconnect attempt */
+ request->rq_timeout = INITIAL_CONNECT_TIMEOUT;
+ lustre_msg_set_timeout(request->rq_reqmsg, request->rq_timeout);
+
+ request->rq_no_resend = request->rq_no_delay = 1;
+ request->rq_send_state = LUSTRE_IMP_CONNECTING;
+ /* Allow a slightly larger reply for future growth compatibility */
+ req_capsule_set_size(&request->rq_pill, &RMF_CONNECT_DATA, RCL_SERVER,
+ sizeof(struct obd_connect_data)+16*sizeof(__u64));
+ ptlrpc_request_set_replen(request);
+ request->rq_interpret_reply = ptlrpc_connect_interpret;
+
+ CLASSERT(sizeof(*aa) <= sizeof(request->rq_async_args));
+ aa = ptlrpc_req_async_args(request);
+ memset(aa, 0, sizeof *aa);
+
+ aa->pcaa_peer_committed = committed_before_reconnect;
+ aa->pcaa_initial_connect = initial_connect;
+
+ if (aa->pcaa_initial_connect) {
spin_lock(&imp->imp_lock);
imp->imp_replayable = 1;
spin_unlock(&imp->imp_lock);
- lustre_msg_add_op_flags(request->rq_reqmsg,
- MSG_CONNECT_INITIAL);
- }
+ lustre_msg_add_op_flags(request->rq_reqmsg,
+ MSG_CONNECT_INITIAL);
+ }
- if (set_transno)
- lustre_msg_add_op_flags(request->rq_reqmsg,
- MSG_CONNECT_TRANSNO);
+ if (set_transno)
+ lustre_msg_add_op_flags(request->rq_reqmsg,
+ MSG_CONNECT_TRANSNO);
DEBUG_REQ(D_RPCTRACE, request, "(re)connect request (timeout %d)",
request->rq_timeout);
!(ocd->ocd_connect_flags & OBD_CONNECT_IBITS)) {
LCONSOLE_WARN("%s: MDS %s does not support ibits "
"lock, either very old or invalid: "
- "requested "LPX64", replied "LPX64"\n",
+ "requested %#llx, replied %#llx\n",
imp->imp_obd->obd_name,
imp->imp_connection->c_remote_uuid.uuid,
imp->imp_connect_flags_orig,
* disable lru_resize, etc. */
if (old_connect_flags != exp_connect_flags(exp) || init_connect) {
CDEBUG(D_HA, "%s: Resetting ns_connect_flags to server "
- "flags: "LPX64"\n", imp->imp_obd->obd_name,
+ "flags: %#llx\n", imp->imp_obd->obd_name,
ocd->ocd_connect_flags);
imp->imp_obd->obd_namespace->ns_connect_flags =
ocd->ocd_connect_flags;
ptlrpc_maybe_ping_import_soon(imp);
GOTO(out, rc);
}
+
+ /* LU-7558: indicate that we are interpretting connect reply,
+ * pltrpc_connect_import() will not try to reconnect until
+ * interpret will finish. */
+ imp->imp_connected = 1;
spin_unlock(&imp->imp_lock);
- LASSERT(imp->imp_conn_current);
+ LASSERT(imp->imp_conn_current);
- msg_flags = lustre_msg_get_op_flags(request->rq_repmsg);
+ msg_flags = lustre_msg_get_op_flags(request->rq_repmsg);
ret = req_capsule_get_size(&request->rq_pill, &RMF_CONNECT_DATA,
RCL_SERVER);
spin_unlock(&imp->imp_lock);
+ if (!exp) {
+ /* This could happen if export is cleaned during the
+ connect attempt */
+ CERROR("%s: missing export after connect\n",
+ imp->imp_obd->obd_name);
+ GOTO(out, rc = -ENODEV);
+ }
+
/* check that server granted subset of flags we asked for. */
if ((ocd->ocd_connect_flags & imp->imp_connect_flags_orig) !=
ocd->ocd_connect_flags) {
- CERROR("%s: Server didn't granted asked subset of flags: "
- "asked="LPX64" grranted="LPX64"\n",
- imp->imp_obd->obd_name,imp->imp_connect_flags_orig,
+ CERROR("%s: Server didn't grant requested subset of flags: "
+ "asked=%#llx granted=%#llx\n",
+ imp->imp_obd->obd_name, imp->imp_connect_flags_orig,
ocd->ocd_connect_flags);
GOTO(out, rc = -EPROTO);
}
+ if ((ocd->ocd_connect_flags2 & imp->imp_connect_flags2_orig) !=
+ ocd->ocd_connect_flags2) {
+ CERROR("%s: Server didn't grant requested subset of flags2: "
+ "asked=%#llx granted=%#llx\n",
+ imp->imp_obd->obd_name, imp->imp_connect_flags2_orig,
+ ocd->ocd_connect_flags2);
+ GOTO(out, rc = -EPROTO);
+ }
+
if (!(imp->imp_connect_flags_orig & OBD_CONNECT_LIGHTWEIGHT) &&
(imp->imp_connect_flags_orig & OBD_CONNECT_MDS_MDS) &&
(imp->imp_connect_flags_orig & OBD_CONNECT_FID) &&
}
}
- if (!exp) {
- /* This could happen if export is cleaned during the
- connect attempt */
- CERROR("%s: missing export after connect\n",
- imp->imp_obd->obd_name);
- GOTO(out, rc = -ENODEV);
- }
old_connect_flags = exp_connect_flags(exp);
exp->exp_connect_data = *ocd;
imp->imp_obd->obd_self_export->exp_connect_data = *ocd;
if (!memcmp(&old_hdl, lustre_msg_get_handle(request->rq_repmsg),
sizeof (old_hdl))) {
LCONSOLE_WARN("Reconnect to %s (at @%s) failed due "
- "bad handle "LPX64"\n",
+ "bad handle %#llx\n",
obd2cli_tgt(imp->imp_obd),
imp->imp_connection->c_remote_uuid.uuid,
imp->imp_dlm_handle.cookie);
* with server again */
if ((MSG_CONNECT_RECOVERING & msg_flags)) {
CDEBUG(level,"%s@%s changed server handle from "
- LPX64" to "LPX64
+ "%#llx to %#llx"
" but is still in recovery\n",
obd2cli_tgt(imp->imp_obd),
imp->imp_connection->c_remote_uuid.uuid,
} else {
LCONSOLE_WARN("Evicted from %s (at %s) "
"after server handle changed from "
- LPX64" to "LPX64"\n",
+ "%#llx to %#llx\n",
obd2cli_tgt(imp->imp_obd),
imp->imp_connection-> \
c_remote_uuid.uuid,
imp->imp_remote_handle =
*lustre_msg_get_handle(request->rq_repmsg);
imp->imp_last_replay_transno = 0;
+ imp->imp_replay_cursor = &imp->imp_committed_list;
IMPORT_SET_STATE(imp, LUSTRE_IMP_REPLAY);
} else {
DEBUG_REQ(D_HA, request, "%s: evicting (reconnect/recover flags"
if (lustre_msg_get_last_committed(request->rq_repmsg) > 0 &&
lustre_msg_get_last_committed(request->rq_repmsg) <
aa->pcaa_peer_committed) {
- CERROR("%s went back in time (transno "LPD64
- " was previously committed, server now claims "LPD64
+ CERROR("%s went back in time (transno %lld"
+ " was previously committed, server now claims %lld"
")! See https://bugzilla.lustre.org/show_bug.cgi?"
"id=9646\n",
obd2cli_tgt(imp->imp_obd), aa->pcaa_peer_committed,
obd2cli_tgt(imp->imp_obd),
imp->imp_connection->c_remote_uuid.uuid);
ptlrpc_connect_import(imp);
+ spin_lock(&imp->imp_lock);
+ imp->imp_connected = 0;
imp->imp_connect_tried = 1;
+ spin_unlock(&imp->imp_lock);
RETURN(0);
}
out:
+ spin_lock(&imp->imp_lock);
+ imp->imp_connected = 0;
imp->imp_connect_tried = 1;
+ spin_unlock(&imp->imp_lock);
if (rc != 0) {
IMPORT_SET_STATE(imp, LUSTRE_IMP_DISCON);
git://git.whamcloud.com / fs / lustre-release.git / blobdiff