Hi,
I experienced similar problems after upgrading 9.x to 10.x on J4350
(with BGP and flowd as well). The JTAC was helpless with this case.
Since I had 90+ percent of memory utilization, I decided to upgrade RAM
from 1 to 2GB and it solved the issue (CPU load has decreased).
New JUNOS (10.x) consumes much more resources than the previous versions.
BR,
r.
Hi,
I upgraded our J2320 from 9.5 to 10.2 this week - the box is now crawling
with CPU consistently 90%. As far as I can see we're still well within the
spec of the box though - aggregate traffic is< 100Mb/s; no. of flow
sessions is between 10k-20k. It is not doing anything overly-complicated -
firewalling between 5 zones, a handful of bgp sessions exchanging only a
small number of routes. "Top" shows that it is flowd that is consuming the
CPU but that doesn't really tell me anything useful:
show chassis routing-engine
Routing Engine status:
Temperature 35 degrees C / 95 degrees F
CPU temperature 44 degrees C / 111 degrees F
Total memory 512 MB Max 451 MB used ( 88 percent)
Control plane memory 298 MB Max 271 MB used ( 91 percent)
Data plane memory 214 MB Max 180 MB used ( 84 percent)
CPU utilization:
User 9 percent
Real-time threads 91 percent
Kernel 0 percent
Idle 0 percent
Model RE-J2320-2000
Serial ID VH6168
Start time 2011-03-08 22:56:09 GMT
Uptime 14 hours, 40 minutes, 31 seconds
Last reboot reason 0x2:watchdog
Load averages: 1 minute 5 minute 15 minute
1.69 1.60 1.48
show chassis forwarding
FWDD status:
State Online
Microkernel CPU utilization 11 percent
Real-time threads CPU utilization 87 percent
Heap utilization 84 percent
Buffer utilization 0 percent
Uptime: 3 hours, 25 minutes, 2 seconds
Anyone any thoughts before I roll back?
regards
Liam.
_______________________________________________
juniper-nsp mailing list juniper-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/juniper-nsp
_______________________________________________
juniper-nsp mailing list juniper-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/juniper-nsp
