Tatulescu Andrei wrote:
In acest moment prin acest server limitez un trafic de 200 - 300 Mbps si aprox
15 - 20 clase de ip-uri. Problema este ca am foarte mare load-ul pe procesor,
in categoria si (80%-90%), atunci cand sunt on limitarile. Am citit la
lartc.org faptul ca trebuie sa folosesc hash-uri pentru a scadea dramatic
load-ul pe procesor si astfel am trecut la implementarea acestor hash-uri, dar
in momentul in care dau comanda :
tc filter add dev eth0 parent 1: prio 15 u32 ht 800:: match ip dst 0/0 hashkey
mask 0x0000ff00 at 16 link 99:
Asta mi-am scris eu cind am sapat pe tema hashing filter, vezi daca te ajuta,
documentatia de pe net e cam vaga pentru iq-ul meu si mi-a luat un pic pina sa
ma prind de niste chestii. Nu garantez ca e si corecta.
---
Filters can be extended to 256 buckets; you define a rule that maps each
packet to a specific bucket (based for example on 8 bits from src or dst) and
make sure that the rules for that packet are in that bucket and not elsewhere.
Keep in mind that each filter prio creates a separate filter tree, so make
sure not to give different prio's to commands that set up the hashing.
Sample:
# define a hashing filter on a specific prio
tc filter add dev eth0 parent 1:0 handle 2: prio 5 protocol ip u32 divisor 256
# define hashing rule (offsets: 12 = src, 16 = src)
tc filter add dev eth0 parent 1:0 prio 5 protocol ip u32 match ip dst
0.0.0.0/0 hashkey mask 0x000000ff at 16 link 2:
# place a filter in a specific bucket
tc filter add dev eth0 parent 1:0 prio 5 protocol ip u32 ht 2:a: match ip dst
10.0.0.10 flowid 1:4
--
Dan Borlovan
Datagroup-Int
_______________________________________________
RLUG mailing list
[email protected]
http://lists.lug.ro/mailman/listinfo/rlug
