Unable to connect MSA 2052 in remote site

[Solitude2]

I have 2 MSA 2052 deployed in location A and B , location A and B are connected via VPN. We have allowed port 3260 for Isci replication however we are still seeing ( Error code 427 (Communication error) & Ping discovery failed ).  Other than port 3260 are there any other ports required by MSA to establish connectivity?

[ArunKKR]

Hi,

From MSA replication guide:

https://www.hpe.com/psnow/doc/4aa1-0977enw?jumpid=in_lit-psnow-red

Network requirements
The following is a guideline for setting up iSCSI controllers with 1 Gb and 10 Gb ports to use with Remote Snap. The two arrays do not have to be in the same subnet, but must be connected to a network whose route tables and firewall or port-blocking settings allow iSCSI traffic (port 3260) to pass between the two systems. CHAP must be configured appropriately on both arrays. If jumbo frames are enabled on either array, it must be enabled on the other array and all network devices in the path between the arrays.

It does not list any other ports.

Have you tried using query peer-connection command to check the results?

[Solitude2]

Thanks for the link  i was able to get the connection up the issue was with the crendentials. But now i have another issue on Site A under replications i am see A1B1 under remote ports ( 2 10G are connected) but on the Site B i only see B1 under local ports Interface A1 is up though but for some reason its not showing under local ports any pointers why or any specific logs to check?  The only difference is that in site A both the 10G are connected to single switch where as in Site B the 10G's are connected to a switch clusters for redundancy.

[ArunKKR]

Hi,

Could you check whether the output of show peer-connections on both the MSAs are identical(except for remote and local port entry)?

Sample output from primary and secondary arrays:

# show peer-connections
Peer Connections
----------------
Peer Connection Name: Peer
Peer Connection Type: iSCSI
Connection Status: Online
Health: OK
Health Reason:
Health Recommendation:

  Local Port     Port Address
  ----------------------------------------------------------
  A1             10.55.202.202
  B1             10.55.202.205

  Remote Port    Port Address
  ----------------------------------------------------------
  A3             10.55.203.201
  B3             10.55.203.202

Success: Command completed successfully. (2021-07-19 17:13:09)

# show peer-connections
Peer Connections
----------------
Peer Connection Name: Peer
Peer Connection Type: iSCSI
Connection Status: Online
Health: OK
Health Reason:
Health Recommendation:

  Local Port     Port Address
  ----------------------------------------------------------
  A3             10.55.203.201
  B3             10.55.203.202

  Remote Port    Port Address
  ----------------------------------------------------------
  A1             10.55.202.202
  B1             10.55.202.205

Success: Command completed successfully. (2021-07-19 15:02:08)
#

[Solitude2]

Here's the output from both the arrays. The arrays are in 2 different locatons . Only A1 shows as connected whereas B1 is also connected and up but does not show under the peer connections 

Here's the output from the arrays whose B1  fails to appear as local port and connected port.

[ArunKKR]

Hi,

The only reason that I could think of is A1 B1 IPs at remote site not able to communicate with each other.

Is there any vlan or network restrictions that stop A1 and B1 IPs from communicating with each other?

[Solitude2]

Are you refering to A1B1 at Site A where only A1 is visible ? If yes then A1 and B1 in Site A are in same vlan and there is no restriction. Site B's setup is similar as well except on Site B A1B1 are connected to the same switch where as in Site A A1B1 are connected to  different interfaces on 2 differnt switches which are in stack..

[ArunKKR]

Hi,

I am referring to site A where only A1 is visible under show peer-connections.

You might need to log a support case to get the logs reviewed if you are sure that A1 B1 IPs in site A can communicate with each other.

Try restarting storage controllers A and B one by one during a period of low IOs to check whether it resolves the issue before logging a support case.