LU-9121 lnet: select best peer and local net

Select the healthiest and highest priority peer and local net when
sending a message.

Lustre-change: https://review.whamcloud.com/34352
Lustre-commit: dff6587805ddad212ab48e5bedacbc7846542b7b

Test-Parameters: trivial testlist=lnet-selftest,sanity-lnet
Signed-off-by: Amir Shehata <ashehata@whamcloud.com>
Change-Id: I42717e7fdc3226c6faa7c59c713f18422e27f2e5
Reviewed-on: https://review.whamcloud.com/c/ex/lustre-release/+/52444
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Frank Sehr <fsehr@whamcloud.com>

diff --git a/lnet/include/lnet/lib-lnet.h b/lnet/include/lnet/lib-lnet.h
index f681088..68793d2 100644 (file)
--- a/lnet/include/lnet/lib-lnet.h
+++ b/lnet/include/lnet/lib-lnet.h
@@ -575,6 +575,7 @@ int lnet_get_rtr_pool_cfg(int idx, struct lnet_ioctl_pool_cfg *pool_cfg);
 struct lnet_ni *lnet_get_next_ni_locked(struct lnet_net *mynet,
                                        struct lnet_ni *prev);
 struct lnet_ni *lnet_get_ni_idx_locked(int idx);
+int lnet_get_net_healthv_locked(struct lnet_net *net);
 
 extern int libcfs_ioctl_getdata(struct libcfs_ioctl_hdr **hdr_pp,
                                struct libcfs_ioctl_hdr __user *uparam);

diff --git a/lnet/include/lnet/lib-types.h b/lnet/include/lnet/lib-types.h
index d30a2cf..fac6be2 100644 (file)
--- a/lnet/include/lnet/lib-types.h
+++ b/lnet/include/lnet/lib-types.h
@@ -390,6 +390,9 @@ struct lnet_net {
         * lnet/include/lnet/nidstr.h */
        __u32                   net_id;
 
+       /* round robin selection */
+       __u32                   net_seq;
+
        /* total number of CPTs in the array */
        __u32                   net_ncpts;
 

diff --git a/lnet/lnet/api-ni.c b/lnet/lnet/api-ni.c
index 0a48d34..4d4f9ae 100644 (file)
--- a/lnet/lnet/api-ni.c
+++ b/lnet/lnet/api-ni.c
@@ -3003,6 +3003,21 @@ lnet_get_ni_idx_locked(int idx)
        return NULL;
 }
 
+int lnet_get_net_healthv_locked(struct lnet_net *net)
+{
+       struct lnet_ni *ni;
+       int best_healthv = 0;
+       int healthv;
+
+       list_for_each_entry(ni, &net->net_ni_list, ni_netlist) {
+               healthv = atomic_read(&ni->ni_healthv);
+               if (healthv > best_healthv)
+                       best_healthv = healthv;
+       }
+
+       return best_healthv;
+}
+
 struct lnet_ni *
 lnet_get_next_ni_locked(struct lnet_net *mynet, struct lnet_ni *prev)
 {

diff --git a/lnet/lnet/lib-move.c b/lnet/lnet/lib-move.c
index fcfc9dd..cd88279 100644 (file)
--- a/lnet/lnet/lib-move.c
+++ b/lnet/lnet/lib-move.c
@@ -1814,11 +1814,22 @@ lnet_handle_send(struct lnet_send_data *sd)
        __u32 routing = send_case & REMOTE_DST;
         struct lnet_rsp_tracker *rspt;
 
-       /*
-        * Increment sequence number of the selected peer so that we
-        * pick the next one in Round Robin.
+       /* Increment sequence number of the selected peer, peer net,
+        * local ni and local net so that we pick the next ones
+        * in Round Robin.
         */
        best_lpni->lpni_seq++;
+       best_lpni->lpni_peer_net->lpn_seq++;
+       best_ni->ni_seq++;
+       best_ni->ni_net->net_seq++;
+
+       CDEBUG(D_NET, "%s NI seq info: [%d:%d:%d] %s LPNI seq info [%d:%d:%d]\n",
+              libcfs_nid2str(best_ni->ni_nid),
+              best_ni->ni_seq, best_ni->ni_net->net_seq,
+              atomic_read(&best_ni->ni_tx_credits),
+              libcfs_nid2str(best_lpni->lpni_nid),
+              best_lpni->lpni_seq, best_lpni->lpni_peer_net->lpn_seq,
+              best_lpni->lpni_txcredits);
 
        /*
         * grab a reference on the peer_ni so it sticks around even if
@@ -2013,8 +2024,7 @@ lnet_find_best_ni_on_spec_net(struct lnet_ni *cur_best_ni,
                              struct lnet_peer *peer,
                              struct lnet_peer_net *peer_net,
                              struct lnet_msg *msg,
-                             int cpt,
-                             bool incr_seq)
+                             int cpt)
 {
        struct lnet_net *local_net;
        struct lnet_ni *best_ni;
@@ -2034,9 +2044,6 @@ lnet_find_best_ni_on_spec_net(struct lnet_ni *cur_best_ni,
        best_ni = lnet_get_best_ni(local_net, cur_best_ni,
                                   peer, peer_net, msg, cpt);
 
-       if (incr_seq && best_ni)
-               best_ni->ni_seq++;
-
        return best_ni;
 }
 
@@ -2273,8 +2280,7 @@ lnet_handle_find_routed_path(struct lnet_send_data *sd,
                lpn = gwni->lpni_peer_net;
                sd->sd_best_ni = lnet_find_best_ni_on_spec_net(NULL, gw, lpn,
                                                               sd->sd_msg,
-                                                              sd->sd_md_cpt,
-                                                              true);
+                                                              sd->sd_md_cpt);
                if (!sd->sd_best_ni) {
                        CERROR("Internal Error. Expected local ni on %s but non found: %s\n",
                               libcfs_net2str(lpn->lpn_net_id),
@@ -2359,9 +2365,15 @@ struct lnet_ni *
 lnet_find_best_ni_on_local_net(struct lnet_peer *peer, struct lnet_msg *msg,
                               int md_cpt, bool discovery)
 {
-       struct lnet_peer_net *peer_net = NULL;
+       struct lnet_peer_net *lpn = NULL;
+       struct lnet_peer_net *best_lpn = NULL;
+       struct lnet_net *net = NULL;
+       struct lnet_net *best_net = NULL;
        struct lnet_ni *best_ni = NULL;
-       int lpn_healthv = 0;
+       int best_lpn_healthv = 0;
+       int best_net_healthv = 0;
+       int net_healthv;
+       bool exit = false;
 
        /*
         * The peer can have multiple interfaces, some of them can be on
@@ -2371,39 +2383,68 @@ lnet_find_best_ni_on_local_net(struct lnet_peer *peer, struct lnet_msg *msg,
         */
 
        /* go through all the peer nets and find the best_ni */
-       list_for_each_entry(peer_net, &peer->lp_peer_nets, lpn_peer_nets) {
+       list_for_each_entry(lpn, &peer->lp_peer_nets, lpn_peer_nets) {
                /*
                 * The peer's list of nets can contain non-local nets. We
                 * want to only examine the local ones.
                 */
-               if (!lnet_get_net_locked(peer_net->lpn_net_id))
+               net = lnet_get_net_locked(lpn->lpn_net_id);
+               if (!net)
                        continue;
 
-               /* always select the lpn with the best health */
-               if (lpn_healthv <= peer_net->lpn_healthv)
-                       lpn_healthv = peer_net->lpn_healthv;
-               else
-                       continue;
-
-               best_ni = lnet_find_best_ni_on_spec_net(best_ni, peer, peer_net,
-                                                       msg, md_cpt, false);
+               net_healthv = lnet_get_net_healthv_locked(net);
 
                /*
                 * if this is a discovery message and lp_disc_net_id is
                 * specified then use that net to send the discovery on.
                 */
-               if (peer->lp_disc_net_id == peer_net->lpn_net_id &&
-                   discovery)
-                       break;
+               if (peer->lp_disc_net_id == lpn->lpn_net_id &&
+                   discovery) {
+                       exit = true;
+                       goto select_lpn;
+               }
+
+               if (!best_lpn)
+                       goto select_lpn;
+
+               /* always select the lpn with the best health */
+               if (best_lpn_healthv > lpn->lpn_healthv)
+                       continue;
+               else if (best_lpn_healthv < lpn->lpn_healthv)
+                       goto select_lpn;
+
+               /* select the preferred peer and local nets */
+               if (best_net_healthv > net_healthv)
+                       continue;
+               else if (best_net_healthv < net_healthv)
+                       goto select_lpn;
+
+               if (best_lpn->lpn_seq < lpn->lpn_seq)
+                       continue;
+               else if (best_lpn->lpn_seq > lpn->lpn_seq)
+                       goto select_lpn;
 
-               best_ni = lnet_find_best_ni_on_spec_net(best_ni, peer,
-                                                  peer_net, msg, md_cpt,
-                                                  false);
+               /* round robin over the local networks */
+               if (best_net->net_seq <= net->net_seq)
+                       continue;
+
+select_lpn:
+               best_net_healthv = net_healthv;
+               best_lpn_healthv = lpn->lpn_healthv;
+               best_lpn = lpn;
+               best_net = net;
+
+               if (exit)
+                       break;
        }
 
-       if (best_ni)
-               /* increment sequence number so we can round robin */
-               best_ni->ni_seq++;
+       if (best_lpn) {
+               /* Select the best NI on the same net as best_lpn chosen
+                * above
+                */
+               best_ni = lnet_find_best_ni_on_spec_net(NULL, peer,
+                                                       best_lpn, msg, md_cpt);
+       }
 
        return best_ni;
 }
@@ -2443,6 +2484,7 @@ static int
 lnet_select_preferred_best_ni(struct lnet_send_data *sd)
 {
        struct lnet_ni *best_ni = NULL;
+       struct lnet_peer_ni *best_lpni = sd->sd_best_lpni;
 
        /*
         * We must use a consistent source address when sending to a
@@ -2462,19 +2504,20 @@ lnet_select_preferred_best_ni(struct lnet_send_data *sd)
                best_ni = lnet_find_existing_preferred_best_ni(sd->sd_best_lpni,
                                                               sd->sd_cpt);
 
-       if (!best_ni)
-               best_ni = lnet_find_best_ni_on_spec_net(NULL, sd->sd_peer,
-                                               sd->sd_best_lpni->lpni_peer_net,
-                                               sd->sd_msg,
-                                               sd->sd_md_cpt,
-                                               true);
-
        /* If there is no best_ni we don't have a route */
        if (!best_ni) {
-               CERROR("no path to %s from net %s\n",
-                       libcfs_nid2str(sd->sd_best_lpni->lpni_nid),
-                       libcfs_net2str(sd->sd_best_lpni->lpni_net->net_id));
-               return -EHOSTUNREACH;
+               best_ni =
+                 lnet_find_best_ni_on_spec_net(NULL, sd->sd_peer,
+                                               sd->sd_best_lpni->lpni_peer_net,
+                                               sd->sd_msg,
+                                               sd->sd_md_cpt);
+               /* If there is no best_ni we don't have a route */
+               if (!best_ni) {
+                       CERROR("no path to %s from net %s\n",
+                               libcfs_nid2str(best_lpni->lpni_nid),
+                               libcfs_net2str(best_lpni->lpni_net->net_id));
+                       return -EHOSTUNREACH;
+               }
        }
 
        sd->sd_best_ni = best_ni;
@@ -2524,8 +2567,7 @@ lnet_handle_any_local_nmr_dst(struct lnet_send_data *sd)
                                                               sd->sd_peer,
                                                               sd->sd_best_lpni->lpni_peer_net,
                                                               sd->sd_msg,
-                                                              sd->sd_md_cpt,
-                                                              true);
+                                                              sd->sd_md_cpt);
                if (!sd->sd_best_ni) {
                        CERROR("Unable to forward message to %s. No local NI available\n",
                               libcfs_nid2str(sd->sd_dst_nid));
@@ -2558,8 +2600,7 @@ lnet_handle_any_mr_dsta(struct lnet_send_data *sd)
                sd->sd_best_ni =
                  lnet_find_best_ni_on_spec_net(NULL, sd->sd_peer,
                                                sd->sd_best_lpni->lpni_peer_net,
-                                               sd->sd_msg, sd->sd_md_cpt,
-                                               true);
+                                               sd->sd_msg, sd->sd_md_cpt);
 
                if (!sd->sd_best_ni) {
                        /*