Reliable data storage for decentralized computer systems

Patent metadata

Field	Value
Publication number	US-2018255138-A1
Application number	US-201715450904-A
Country	US
Kind code	A1
Filing date	Mar 6, 2017
Priority date	Mar 6, 2017
Publication date	Sep 6, 2018
Grant date	—

Abstract

A method includes receiving a data object for storage in a storage system. The storage system includes a number of datacenters (s) interconnected by a first network. Each of the datacenters is located in a geographic location that is different than any geographic locations of any other of the datacenters. The method includes creating secondary copies of the data object. A number of secondary copies is equal to at least s−1. The method includes, in accordance with a placement map of at least one of the datacenters, storing a primary copy of the data object in one of the datacenters. The method also includes, in each other of the datacenters, storing at least one of the secondary copies. The method also includes monitoring, via a plurality of data monitors, an accessibility of data stored in the storage system. The data includes the primary copy and the secondary copies of the data objects.

First claim

What is claimed:

1 . A method comprising:
identifying a datacenter group having a number of datacenters (s), wherein each datacenter of the s datacenters comprises a plurality of data monitors, wherein a number of the plurality of data monitors in each datacenter is equal to at least s+1; for each datacenter of the datacenter group, creating a placement map; receiving a data object to be stored in the datacenter group; creating secondary copies of the data object, wherein a number of secondary copies is equal to at least s−1; and in accordance with the placement maps, storing a primary copy of the data object in one of the datacenters and, in each other of the datacenters, storing at least one of the secondary copies.

2 . The method of claim 1 , wherein the datacenters of the datacenter group are interconnected via low-latency network connections.

3 . The method of claim 3 , wherein the low-latency network connections comprise a virtual private network (VPN) tunnel.

4 . The method of claim 1 , further comprising monitoring, via the plurality of data monitors, for a change or loss of data within each of the datacenters.

5 . The method of claim 1 , wherein at least one of the datacenters comprises a lightweight data center.

6 . The method of claim 1 , further comprising:
receiving a query for the data object; determining that the primary copy of the data object is inaccessible; based on the placement map, identifying a location of at least one of the secondary copies of the data object; and replying to the query based at least on the at least one of the secondary copies.

7 . The method of claim 6 , further comprising:
retrieving the at least one of the secondary copies through a VPN tunnel.

8 . A system comprising:
a first network; and a number of datacenters (s) interconnected via the first network; a data controller communicatively coupled to the first network, the data controller comprising:
an input/output for communicating via a second network;
a processor communicatively coupled to the input/output; and
memory storing instructions that cause the processor to effectuate operations comprising:
for each datacenter of the s datacenters, creating a placement map;
receiving a data object via the second network;
creating secondary copies of the data object, wherein a number of secondary copies is equal to at least s−1; and
in accordance with the placement maps, storing a primary copy of the data object in one of the datacenters and, in each of the other datacenters, storing at least one of the secondary copies,
wherein each of the s datacenters are located at geographically distinct locations from one another.

9 . The system of claim 8 , wherein the first network comprises a virtual private network (VPN).

10 . The system of claim 8 , wherein the operations further comprise:
receiving, from the second network, a query for the data object; determining that the primary copy of the data object is inaccessible; based on the placement map, identifying a location of at least one of the secondary copies of the data object; and replying to the query based on the at least one of the secondary copies.

11 . The system of claim 8 , wherein the operations further comprise:
receiving, from the second network, a query for the data object; based on the placement map, identifying a location of the primary copy of the data object; retrieving, via the first network, the primary copy from the location; and transmitting, via the second network, a reply to the query based on the primary copy.

12 . The system of claim 8 , further comprising a plurality of data monitors for monitoring the s datacenters, wherein a number of the plurality of data monitors in each datacenter is equal to at least 1+s.

13 . The system of claim 8 , wherein each of the s datacenters stores at least one of the plurality of data monitors.

14 . The system of claim 8 , wherein a latency of the first network is lower than a latency of the second network.

15 . A method comprising:
receiving a data object for storage in a storage system, the storage system comprising a number of datacenters (s) interconnected by a first network; creating secondary copies of the data object, wherein a number of secondary copies is equal to at least s−1; in accordance with a placement map of at least one of the datacenters, storing a primary copy of the data object in one of the datacenters and, in each other of the datacenters, storing at least one of the secondary copies; and monitoring, via a plurality of data monitors, an accessibility of data stored in the storage system, the data including the primary copy and the secondary copies of the data objects.

16 . The method of claim 15 , further comprising:
based at least on the monitoring, detecting, by a monitor of the plurality of monitors, an event affecting the accessibility of the primary copy; determining, by the monitor, that a query for the data object was received via a second network; communicating, by the monitor, to at least one other monitor of the plurality of data monitors, wherein the at least one other monitor is associated with a second datacenter of the datacenters, wherein at least one of the secondary copies of the data object is stored in the second datacenter; receiving, by the monitor, a transmission based on the at least one of the secondary copies from the second datacenter; and transmitting a reply to the query based at least on the transmission.

17 . The method of claim 17 , wherein a latency of the first network is lower than a latency of the second network.

18 . The method of claim 17 , wherein the first network comprises a virtual private network (VPN).

19 . The method of claim 15 , wherein at least one of the data monitors is a Ceph instance.

20 . The method of claim 15 , wherein a number of the data monitors is equal to at least s+1.

Assignees

• At & T Ip I Lp

Inventors

• Hall Robert J

Classifications

• H04L12/4641

  Virtual LANs, VLANs, e.g. virtual private networks [VPN] (LAN interconnection over a bridge based backbone H04L12/462; encapsulation techniques H04L12/4633; routing of packets H04L45/00; packet switches H04L49/00; virtual private networks for security H04L63/0272) · CPC title

• H04L43/08

  Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters · CPC title

• H04L12/4633

  Interconnection of networks using encapsulation techniques, e.g. tunneling · CPC title

• H04L67/1097 Primary

  for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS] · CPC title

• H04L67/1095

  Replication or mirroring of data, e.g. scheduling or transport for data synchronisation between network nodes · CPC title