don;t see this issue on kilo setup , so close it
** Changed in: neutron
Status: Incomplete => Invalid
--
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed to neutron.
https://bugs.launchpad.net/bugs/1498472
Title:
lbaas:after
Public bug reported:
Defect:XB300349
New Note:
State: Open by: liujgwx on 28 August 2015 07:51:41
add setup infor
my setup is openstack RDO juno with docker
Controller
1) nsj13
PW: Lab4man1
Public IP: 158.85.164.68
Private IP: 10.53.87.252
Network Node
1) NSJ17
Public bug reported:
1 create two-arm LB with 1client and 1 backend server on a tenant
2 repeat step1 to create 350 tenants
3 after step 2 , the LB network unstable
log and cpu infor
[root@nsj4 ~]# ip netns exec qlbaas-ade3e627-5fda-4781-855e-25a0f5217d17 ping
20.1.1.100
PING 20.1.1.100
Public bug reported:
LB_NFV[Alchemy] linux kernel soft lockup after create 200 LB on sinalge tenant
with 30M traffic on each LB
CentOS Linux release 7.1.1503 (Core)
reproduce step
1 create 1 client and 1 backend server by docker container
2 create a lb and add backend as member
3 send 30M
Public bug reported:
[Alchemy][LB_namespaceV1]after 817 tenants was create for namespace 1v1
maping, the new LB pool with pending_creat status
test step
1 create a tenant
2 create a pool network with subnet 10.1.1.0
create a vip network with subnet 20.1.1.0
3 create a router and add 2
Public bug reported:
LB_NFV[Alchemy]iperf3 traffic was stoped after send traffice about
10-15s
root@instance-0002:/# iperf3 -c 20.1.1.100 -t 1000 -i 1 -p 5001
Connecting to host 20.1.1.100, port 5001
[ 4] local 20.1.1.200 port 59473 connected to 20.1.1.100 port 5001
[ ID] Interval
Public bug reported:
LB_NFV[Alchemy]LB performace just have 1 G low peformance than LB
bypass have 4G
setup infor
for LB bypass , client directly send traffic to server without LB , we have 4G
performance
and for LB, client send traffic with LB , we just have 1G traffic
so LB is a bottleneck
7 matches
Mail list logo
