From 6c12c8ab7629d5b1f11595bf779a1ebb5ba09746 Mon Sep 17 00:00:00 2001 From: Donald Sharp Date: Thu, 31 Mar 2016 08:20:53 -0400 Subject: [PATCH] zebra: Fix Startup with > 1k interfaces Zebra in rt_netlink.c has a while (1) loop that handles recvmsg from the netlink socket. In early bootup a situation can occur whereby the netlink messages take a long time to parse. This time starts to take an exponential amount of time the more netlink messages that you read in. There reaches a point where the incoming netlink messages are coming in at about the same rate that they are processed. This ends up causing the while (1) loop to never exit. Eventually this causes quagga to fail when the watchdog message is never sent to systemd. This patch attempts to address this deficiency in that we allow for a pause from reading in netlink messages to allow other work to be done. This pause drains the work queue items created by the netlink received data and allows zebra to respond to other system input. I believe we will need to come back in and modify zebra startup a bit more. There are ineffiencies that need to be addressed as part of boot up. Ticket: CM-9992 Signed-off-by: Donald Sharp Reviewed-by: Don Slice Reviewed-by: Wilson Kok --- zebra/rt_netlink.c | 23 ++++++++++++++--------- 1 file changed, 14 insertions(+), 9 deletions(-) diff --git a/zebra/rt_netlink.c b/zebra/rt_netlink.c index b95ef8462b..5686d8e528 100644 --- a/zebra/rt_netlink.c +++ b/zebra/rt_netlink.c @@ -290,11 +290,12 @@ vrf_lookup_by_table (u_int32_t table_id) static int netlink_parse_info (int (*filter) (struct sockaddr_nl *, struct nlmsghdr *, ns_id_t), - struct nlsock *nl, struct zebra_ns *zns) + struct nlsock *nl, struct zebra_ns *zns, int count) { int status; int ret = 0; int error; + int read_in = 0; while (1) { @@ -312,6 +313,9 @@ netlink_parse_info (int (*filter) (struct sockaddr_nl *, struct nlmsghdr *, }; struct nlmsghdr *h; + if (count && read_in >= count) + return 0; + status = recvmsg (nl->sock, &msg, 0); if (status < 0) { @@ -336,7 +340,8 @@ netlink_parse_info (int (*filter) (struct sockaddr_nl *, struct nlmsghdr *, nl->name, msg.msg_namelen); return -1; } - + + read_in++; for (h = (struct nlmsghdr *) buf; NLMSG_OK (h, (unsigned int) status); h = NLMSG_NEXT (h, status)) { @@ -1464,7 +1469,7 @@ interface_lookup_netlink (struct zebra_ns *zns) ret = netlink_request (AF_PACKET, RTM_GETLINK, &zns->netlink_cmd); if (ret < 0) return ret; - ret = netlink_parse_info (netlink_interface, &zns->netlink_cmd, zns); + ret = netlink_parse_info (netlink_interface, &zns->netlink_cmd, zns, 0); if (ret < 0) return ret; @@ -1472,7 +1477,7 @@ interface_lookup_netlink (struct zebra_ns *zns) ret = netlink_request (AF_INET, RTM_GETADDR, &zns->netlink_cmd); if (ret < 0) return ret; - ret = netlink_parse_info (netlink_interface_addr, &zns->netlink_cmd, zns); + ret = netlink_parse_info (netlink_interface_addr, &zns->netlink_cmd, zns, 0); if (ret < 0) return ret; @@ -1481,7 +1486,7 @@ interface_lookup_netlink (struct zebra_ns *zns) ret = netlink_request (AF_INET6, RTM_GETADDR, &zns->netlink_cmd); if (ret < 0) return ret; - ret = netlink_parse_info (netlink_interface_addr, &zns->netlink_cmd, zns); + ret = netlink_parse_info (netlink_interface_addr, &zns->netlink_cmd, zns, 0); if (ret < 0) return ret; #endif /* HAVE_IPV6 */ @@ -1500,7 +1505,7 @@ netlink_route_read (struct zebra_ns *zns) ret = netlink_request (AF_INET, RTM_GETROUTE, &zns->netlink_cmd); if (ret < 0) return ret; - ret = netlink_parse_info (netlink_routing_table, &zns->netlink_cmd, zns); + ret = netlink_parse_info (netlink_routing_table, &zns->netlink_cmd, zns, 0); if (ret < 0) return ret; @@ -1509,7 +1514,7 @@ netlink_route_read (struct zebra_ns *zns) ret = netlink_request (AF_INET6, RTM_GETROUTE, &zns->netlink_cmd); if (ret < 0) return ret; - ret = netlink_parse_info (netlink_routing_table, &zns->netlink_cmd, zns); + ret = netlink_parse_info (netlink_routing_table, &zns->netlink_cmd, zns, 0); if (ret < 0) return ret; #endif /* HAVE_IPV6 */ @@ -1642,7 +1647,7 @@ netlink_talk (struct nlmsghdr *n, struct nlsock *nl, struct zebra_ns *zns) * Get reply from netlink socket. * The reply should either be an acknowlegement or an error. */ - return netlink_parse_info (netlink_talk_filter, nl, zns); + return netlink_parse_info (netlink_talk_filter, nl, zns, 0); } /* Routing table change via netlink interface. */ @@ -2471,7 +2476,7 @@ static int kernel_read (struct thread *thread) { struct zebra_ns *zns = (struct zebra_ns *)THREAD_ARG (thread); - netlink_parse_info (netlink_information_fetch, &zns->netlink, zns); + netlink_parse_info (netlink_information_fetch, &zns->netlink, zns, 5); zns->t_netlink = thread_add_read (zebrad.master, kernel_read, zns, zns->netlink.sock); -- 2.39.5