Добрый день, коллеги.
Поднял сабж. Два lagg (вход/выход), тюнинг, MPD5.5 собирает netflow. По vmstat -z вижу что увеличивается счетчик FAILURES на позиции NetFlow cache:
# vmstat -z
ITEM SIZE LIMIT USED FREE REQUESTS FAILURES
UMA Kegs: 208, 0, 79, 6, 79, 0
UMA Zones: 960, 0, 79, 1, 79, 0
UMA Slabs: 568, 0, 4431, 357, 8119, 0
UMA RCntSlabs: 568, 0, 37339, 6, 37339, 0
UMA Hash: 256, 0, 0, 15, 3, 0
16 Bucket: 152, 0, 189, 11, 189, 0
32 Bucket: 280, 0, 423, 11, 427, 1
64 Bucket: 536, 0, 736, 6, 753, 37
128 Bucket: 1048, 0, 1493, 199, 941336, 536
VM OBJECT: 216, 0, 46314, 3582, 6041003, 0
MAP: 232, 0, 7, 25, 7, 0
KMAP ENTRY: 120, 277605, 259, 1477, 231731, 0
MAP ENTRY: 120, 0, 881, 9318, 13323324, 0
DP fakepg: 120, 0, 0, 0, 0, 0
SG fakepg: 120, 0, 0, 0, 0, 0
mt_zone: 2056, 0, 207, 52, 207, 0
16: 16, 0, 5954, 3958, 10948191, 0
32: 32, 0, 10866, 4587, 1284742692, 0
64: 64, 0, 6297, 5071, 6159039, 0
128: 128, 0, 22193, 9040, 2963360, 0
256: 256, 0, 58051, 8054, 2586309256, 0
512: 512, 0, 4312, 3395, 7714622, 0
1024: 1024, 0, 307, 1105, 247887, 0
2048: 2048, 0, 599, 1101, 2989, 0
4096: 4096, 0, 227, 1388, 402296, 0
Files: 80, 0, 109, 2366, 5055083, 0
TURNSTILE: 136, 0, 3277, 323, 3277, 0
umtx pi: 96, 0, 0, 0, 0, 0
MAC labels: 40, 0, 0, 0, 0, 0
PROC: 1136, 0, 45, 1230, 253630, 0
THREAD: 1120, 0, 1506, 1770, 144196, 0
SLEEPQUEUE: 80, 0, 3277, 522, 3277, 0
VMSPACE: 392, 0, 28, 1372, 289093, 0
cpuset: 72, 0, 57, 343, 196, 0
audit_record: 952, 0, 0, 0, 0, 0
mbuf_packet: 256, 0, 66946, 7422, 2668444338, 0
mbuf: 256, 0, 4, 4093, 3165129665, 0
mbuf_cluster: 2048, 400000, 74368, 6, 74368, 0
mbuf_jumbo_page: 4096, 12800, 0, 152, 166, 0
mbuf_jumbo_9k: 9216, 6400, 0, 0, 0, 0
mbuf_jumbo_16k: 16384, 3200, 0, 0, 0, 0
mbuf_ext_refcnt: 4, 0, 0, 0, 0, 0
NetGraph items: 72, 65540, 0, 2958, 2856406, 0
NetGraph data items: 72, 65540, 1, 2232, 2645441708, 0
ttyinq: 160, 0, 135, 465, 765, 0
ttyoutq: 256, 0, 72, 243, 408, 0
g_bio: 232, 0, 0, 8464, 405343, 0
ata_request: 320, 0, 0, 2868, 101346, 0
ata_composite: 336, 0, 0, 0, 0, 0
VNODE: 472, 0, 95798, 1482, 323900, 0
VNODEPOLL: 112, 0, 1, 65, 1, 0
S VFS Cache: 108, 0, 73923, 28443, 311061, 0
L VFS Cache: 328, 0, 22869, 219, 25611, 0
NAMEI: 1024, 0, 0, 668, 6429193, 0
DIRHASH: 1024, 0, 1930, 374, 2260, 0
pipe: 728, 0, 5, 965, 133616, 0
ksiginfo: 112, 0, 1265, 1738, 6148, 0
itimer: 344, 0, 0, 0, 0, 0
KNOTE: 128, 0, 0, 406, 1873, 0
socket: 680, 25602, 43, 833, 1911099, 0
ipq: 56, 12537, 0, 0, 0, 0
udp_inpcb: 336, 25608, 9, 816, 1780557, 0
udpcb: 16, 25704, 9, 3351, 1780557, 0
tcp_inpcb: 336, 25608, 8, 146, 25, 0
tcpcb: 944, 25600, 8, 88, 25, 0
tcptw: 72, 5150, 0, 200, 5, 0
syncache: 144, 15366, 0, 182, 11, 0
hostcache: 136, 15372, 0, 196, 9, 0
tcpreass: 40, 25032, 0, 0, 0, 0
sackhole: 32, 0, 0, 202, 1, 0
ripcb: 336, 25608, 1, 714, 53945, 0
unpcb: 240, 25600, 15, 625, 76555, 0
rtentry: 200, 0, 4106, 2316, 26342, 0
IPFW dynamic rule: 120, 0, 0, 0, 0, 0
selfd: 56, 0, 2049, 2865, 115271476, 0
SWAPMETA: 288, 116519, 0, 0, 0, 0
Mountpoints: 752, 0, 5, 20, 5, 0
FFS inode: 168, 0, 95764, 1608, 323823, 0
FFS1 dinode: 128, 0, 0, 0, 0, 0
FFS2 dinode: 256, 0, 95764, 1601, 323823, 0
NetFlow cache: 80, 262160, 191175, 70985, 29063801, 939849
В колонке LIMITS находится значение 262160, при большой нагрузке (800 сессий) его не хватает и растет FAILURES. Подскажите, где можно увеличить netflow кэш? Исходники проясняют ситуацию смутно, так как не силен в Си и в архитектуре BSD.
