Unsolved
219 Posts
0
725
DDBoost use only one nic for "restore validate" (NMDA RMAN)
Hi,
we are happy with our DDs for 6 months now. But at this time we have a situation where we get in trouble.
We have a NVE (9.2) configured one DD6300 locally and one DD3300 (6.1.1.20) remotely. On the remote Site we added several large Oracle Database Servers. Backup runs well using both configured nic/ip`s. But when our db stuff do "restore validate" only one nic/ip is using even when they are using more then one process per db.
Config Details:
no LACP
4x 1gbit/s nic eth1a = x.x.x.1 Management (FQDN)+ replication eth1b = x.x.x.2 backup eth1c = x.x.x.3 backup eth1d = x.x.x.4 replication eth1a / d Interface Group replication
eth1b / c Interface Group backup
When a restore is running normally eth1d (replication) is used! Is this true or a bug ?
Is there any option to get better restore time`s instead of adding the extra 2x 10gbit/s Controller ?
Andreas
andreas.scht
219 Posts
0
July 4th, 2019 09:00
Some more Information:
The recover traffic is just running on eth1d (Default Gateway) not like the backup traffic using ifgroup "backup" (ethb/c). Is this because of the default gateway ?
I thought boost will load balancing still for recover using ifgroup "backup". But it seems not.