Reduction of adjacent rack traffic in multi-rack distributed object storage systems
An example system may include a first rack including a first access node and a second rack including a second access node and a storage node. The first and second racks are included within a multi-rack distributed storage system. The first access node is configured to receive over a connection with a client a first request for an object data. The connection with the client is uniquely identified by a connection ID independent of source and destination addresses and ports. The first access node is further configured to determine object data locality details of the object data and relocate the connection to the second access node in the second rack when the object data locality details of the object data indicate the object data is located within the second rack.
1. A system, comprising:
a first rack including a first access node; and
a second rack including a second access node and a storage node,
wherein:
the first rack and the second rack are included within a multi-rack distributed storage system; and
the first access node is configured to:
receive, over a connection with a client, a first request for an object data, wherein:
the connection with the client is uniquely identified by a connection identifier (ID) independent of source and destination addresses and ports; and
the connection ID is associated with a first internet protocol (IP) address for the first access node;
determine object data locality details of the object data; and
relocate, responsive to the object data locality details of the object data indicating that the object data is located within the second rack, the connection to the second access node in the second rack, wherein the connection ID is associated with a second IP address for the second access node after relocation.
2. The system of claim 1 , wherein the first rack includes metadata including the object data locality details for the object data.
3. The system of claim 1 , wherein the first access node is further configured to determine the object data locality details by:
sending a second request to a remote access node in a remote rack within the multi-rack distributed storage system, the remote access node including metadata including the object data locality details of the object data; and
receiving the object data locality details of the object data from the remote access node.
4. The system of claim 3 , wherein the remote rack and the second rack are a same rack within the multi-rack distributed storage system.
5. The system of claim 3 , wherein the remote rack and the second rack are different racks within the multi-rack distributed storage system.
6. The system of claim 1 , wherein the first access node is further configured to relocate the connection by sending the first request from the client for the object data to the second access node in the second rack.
7. The system of claim 6 , wherein the first access node is further configured to relocate the connection by sending the connection ID to the second access node.
8. The system of claim 7 , wherein the second access node is configured to:
send the connection ID to the client; and
send a destination address and a destination port of the second access node to the client.
9. The system of claim 8 , wherein:
the system further includes the client; and
the client is configured to change a destination IP address associated with the connection ID from the first IP address of the first access node in the first rack to the second IP address of the second access node in the second rack.
10. The system of claim 1 , wherein the second access node is configured to send a third request to the storage node in the second rack based on the object data locality details of the object data.
11. The system of claim 10 , wherein the second access node is further configured to return the object data to the client responsive to the first request being a request to get the object data.
12. The system of claim 10 , wherein the second access node is further configured to return a second response to the client responsive to the first request being a request to write the object data to the multi-rack distributed storage system.
13. The system of claim 1 , wherein the connection is based on a Quick User datagram protocol (UDP) Internet Connection (QUIC) protocol.
14. A computer-implemented method, comprising:
receiving, over a connection between a client and a first access node in a first rack, a first request for an object data, wherein:
the connection with the client is uniquely identified by a connection identifier (ID) independent of source and destination addresses and ports; and
the connection ID is associated with a first internet protocol (IP) address for the first access node;
determining, at the first access node, object data locality details of the object data; and
relocating, responsive to the object data locality details of the object data indicating that the object data is located within a second rack, the connection to a second access node in the second rack, wherein:
the connection ID is associated with a second IP address for the second access node after relocation; and
the first and second racks are included within a multi-rack distributed storage system.
15. The computer-implemented method of claim 14 , wherein the first rack includes metadata including the object data locality details for the object data.
16. The computer-implemented method of claim 14 , wherein determining the object data locality details further includes:
sending a second request to a remote access node in a remote rack within the multi-rack distributed storage system, wherein the remote access node includes metadata including the object data locality details of the object data; and
receiving the object data locality details of the object data from the remote access node.
17. The computer-implemented method of claim 14 , wherein relocating the connection further includes sending the first request from the client for the object data to the second access node in the second rack.
18. The computer-implemented method of claim 14 , further comprising sending a third request to a storage node in the second rack based on the object data locality details of the object data.
19. The computer-implemented method of claim 14 , wherein the connection is based on a Quick User datagram protocol (UDP) Internet Connection (QUIC) protocol.
20. A system, comprising:
means for receiving, over a connection between a client and a first access node in a first rack, a first request for an object data, wherein:
the connection with the client is uniquely identified by a connection identifier (ID) independent of source and destination addresses and ports; and
the connection ID is associated with a first internet protocol (IP) address for the first access node;
means for determining, at the first access node, object data locality details of the object data; and
means for relocating, responsive to the object data locality details of the object data indicating that the object data is located within a second rack, the connection to a second access node in the second rack, wherein:
the connection ID is associated with a second IP address for the second access node after relocation; and
the first and second racks are included within a multi-rack distributed storage system.