2015-05-27 05:51 AM
I am confused on something and I was wondering if you could help me out.
I have 2 filers across 2 locations and I have set up a snapvault relationship between the two.
How do I know what interfaces are being used by each side?
The source side I know because I had to define the interface when setting up the SV relationship, but at no point was there an option to specify what the destination interface should be.
The destination side has a number of ethernet ports and 2 of these ports are 10GB LACP. Others are 1GB ports used for management and other stuff.
How do I know for sure that the traffic sent from FilerA is being received on the correct ports for FilerB? Ideally I want the traffic to come down the 2 x 10GB ports on the destination - what must I do on the destination side to achieve this?
Itd be nice if someone could explain the network flow...if possible
2015-05-27 06:03 AM
I'm assuming you have multiple interfaces as per your statement... What's your traceroute to the filer?
On the filer you can run ifstat -a and check the traffic flow of your interface vifs.
Worst comes to worse, you can use host file entries
2015-05-27 06:33 AM
When I ping the destination filer name I get a 172.x.x.x address. I believe this is the management address.
I need to get the traffic to leave the interfaces that make up a 10.22.x.x address and then be received on a 10.22.x.x address on the destination filer.
I have a weird feeling that the traffic is leaving the 10GB interfaces on the source filer (10.22 address) but is destined to go to the 172.x. address and may be being received on the management IP.
When I look at the SV relationship I can see that the destinaation is defined as FILERB:/vol/SV_Snapvault/Q_SV_Snapvault
Am i right in saying that when the snapvault kicks off, it will route to the hostname of the desitnation Filer? In this case, FILERB......which resolved to a 172.x.x.x address?
2015-05-27 06:41 AM
I am just trying to work out when the SV kicks off how does it know what IP/HOST it needs to go to?
Is it going to go to the HOST of the destination filer as defined in the destination path of the SV relationship? If so, then its basically routing to the management interfaces......, whcih isnt good.
2015-05-27 06:52 AM
IP is resolved from hostname you specify in "snapvault start" command. You can use IP address directly here. I am not sure what happens if you give IP alias which is not the same as primary hostname (I know that SnapMirror was rather picky here); but you can try. Using IP address is always an option.
2015-05-27 07:35 AM
How do I check my routing table. Something weird going on with that.
When I ping the hostname of the destination filer I get an address of 172.17.2.94
When I do a traceroute from the source filer to the destination filer, I get this:
traceroute to FILERB (172.17.2.94), 30 hops max, 40 byte packets
1 10.2.2.254 (10.2.2.254) 1.985 ms 2.000 ms 2.998 ms
2 10.1.8.20 (10.1.8.20) 1.038 ms 1.001 ms 0.999 ms
The first hop is the core switch.
The second hop is according to the hosts file on the destination filer the following:
If I do a ping -s arc-filer1, the address that responds is 172.17.2.94
So pinging it (with a minus s resolves the correct address) but tracerouting to it, returns the last hop as 10.1.8.20 which is VIF_NFS........
Any know know what is going on?
Its either being received (on the destination) on the Management address (172.x.x.x) or the wrong VIF (10.1.8.20). I need it to resolve on the 2 x 10GB interfaces, which are on the 10.22.x.x range)
2015-05-27 08:41 AM
You told it to ping 172.17.2.94, why do you expect it to magically start pinging 10.1.8.20 instead? If you want to connect (or ping) 10.1.8.20, just say so. If you want name FILERB on secondary resolve to 10.x.x.x address - add this name to /etc/hosts on secondary with this address.
As to why traffic to 172.17.2.94 goes via 10.x.x.x - I presume that is how your routing table looks like. Please paste "netstat -rn" output on secondary.