# US 2026/0236193 A1 · High Performance Object Storage

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2026/0236193 A1
- Document ID: US20260236193A1
- Status: Published application
- Kind: A1
- Application number: 19/538,335
- Filed: 2026-02-12
- Published: 2026-08-13
- Earliest claimed: 2025-02-13
- Assignees: Google Inc.
- Applicants: Google LLC
- Inventors: Michael James Maloney; Denis Matthew Serenyi; Webb Harrington Horn; Jeffrey Karl Terrace; Christopher James Carlon; Benjamin Chapin Kittridge; IIya Serov; Garrett Ryan Mueller
- Web page: https://patentgrep.com/patent/US20260236193A1

## CPC

- G06F 3/067: {Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]}
- G06F 16/182: Distributed file systems
- G06F 3/0689: {Disk arrays, e.g. RAID, JBOD}
- G06F 16/196: {Specific adaptations of the file system to access devices and non-file objects via standard file system access operations, e.g. pseudo file systems (dedicated interfaces to storage systems G06F3/0601)}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

The present disclosure describes a latency reduction system that provides a number of capabilities to a cloud storage platform. Such capabilities include appendable objects, opening an object session for read, opening an object session for writes, resuming an interrupted read or write object session, reads and writes over object sessions, object fragments, and a program for serving dataplane traffic for the latency reduction system.

## Claims

### Claim 1

A system, comprising:
one or more storage entities configured to store files;
one or more processors in communication with the one or more storage entities, the one or more processors configured to:
receive a request from a client to perform a transaction with respect to an object corresponding to one of the files in the one or more storage entities;
open a connection between the client and the one or more storage entities through the one or more processors, wherein opening the connection comprises establishing and authorizing an object session with metadata services;
receive, from the metadata services, a session token for the object session and an object fragment list for the one or more storage entities;
execute the transaction; and
return a confirmation of the transaction and the session token to the client.

### Claim 2 (depends on 1)

The system of claim 1, wherein the one or more processors are configured to ensure that append requests are honored from only one client at a given time.

### Claim 3 (depends on 2)

The system of claim 2, wherein ensuring that append requests are honored from only one client at a time comprises closing an existing write session from a first client if a subsequent write session is started by a second client.

### Claim 4 (depends on 1)

The system of claim 1, wherein the one or more processors are configured to maintain the connection open for a predetermined period of time during which further append requests can be honored.

### Claim 5 (depends on 1)

The system of claim 1, wherein the one or more processors are configured to automatically close the connection after the predetermined period of time.

### Claim 6 (depends on 1)

The system of claim 1, wherein the one or more processors are configured to reopen, using the session token, a closed connection when further transaction requests are received, without repeating the authorizing with the metadata services.

### Claim 7 (depends on 1)

The system of claim 1, wherein the transaction comprises a write transaction to append data to the object; and wherein the one or more processors are configured to:
determine a portion of the object to which the data should be appended; and
append the data at the determined portion of the object.

### Claim 8 (depends on 7)

The system of claim 7, wherein the determined portion is at a tail end of the object.

### Claim 9 (depends on 7)

The system of claim 7, wherein after appending the data the object is finalized such that no further data can be appended.

### Claim 10 (depends on 9)

The system of claim 9, wherein finalizing the object is performed manually by user request or automatically after a predetermined period of time.

### Claim 11 (depends on 1)

The system of claim 1, wherein the transaction requests are routed to processors physically closest to the physical storage corresponding to the request.

### Claim 12 (depends on 11)

The system of claim 11, wherein the system comprises multiple processors in different locations, and wherein a first of the one or more processors is configured to, upon receiving the transaction request:
determine if the transaction request is directed to a storage location associated with a second processor in a different location than the first processor; and
redirect the transaction request to the second processor.

### Claim 13 (depends on 1)

The system of claim 1, wherein the confirmation of the transaction comprises an acknowledgement of a successful write or data requested in a read transaction.

### Claim 14

A method, comprising:
receiving, at one or more processors in communication with one or more storage entities storing files, a request from a client to perform a transaction with respect to an object corresponding to one of the files in the one or more storage entities;
opening a connection between the client and the one or more storage entities through the one or more processors, wherein opening the connection comprises establishing and authorizing an object session with metadata services;
receiving, from the metadata services, a session token for the object session and an object fragment list for the one or more storage entities;
executing the transaction; and
returning a confirmation of the transaction and the session token to the client.

### Claim 15 (depends on 14)

The method of claim 14, comprising honoring append requests from only one client at a given time by closing an existing write session from a first client if a subsequent write session is started by a second client.

### Claim 16 (depends on 14)

The method of claim 14, comprising maintaining the connection open for a predetermined period of time during which further append requests can be honored, and automatically closing the connection after the predetermined period of time.

### Claim 17 (depends on 14)

The method of claim 14, wherein the transaction comprises a write transaction to append data to the object, the method further comprising:
determining a portion of the object to which the data should be appended; and
appending the data at the determined portion of the object.

### Claim 18 (depends on 17)

The method of claim 17, wherein the determined portion is at a tail end of the object.

### Claim 19 (depends on 17)

The method of claim 17, further comprising finalizing the object after appending the data such that no further data can be appended.

### Claim 20

A non-transitory computer-readable medium storing instructions executable by one or more processors for performing a method, comprising:
receiving, at one or more processors in communication with one or more storage entities storing files, a request from a client to perform a transaction with respect to an object corresponding to one of the files in the one or more storage entities;
opening a connection between the client and the one or more storage entities through the one or more processors, wherein opening the connection comprises establishing and authorizing an object session with metadata services;
receiving, from the metadata services, a session token for the object session and an object fragment list for the one or more storage entities;
executing the transaction; and
returning a confirmation of the transaction and the session token to the client.
