Method and system for preventing data loss within disk-array pairs supporting mirrored logical units

• US 7,058,850 B2
• Filed: 07/31/2002
• Issued: 06/06/2006
• Est. Priority Date: 07/31/2002
• Status: Expired due to Fees

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Cochran, Robert A.
• Primary Examiner(s): Beausoliel, Robert
• Assistant Examiner(s): MCCARTHY, CHRISTOPHER S
• Application Number:
  US10/210,368
• Publication Number:
  US 20040078638A1
• Time in Patent Office: 1,406 Days
• Field of Search: 714/6, 714/43, 714/4, 714/42
• US Class Current: 714/6.12
• CPC Class Codes: G06F 11/1443 Transmit or communication errors
• CPC Class Codes: G06F 11/2058 using more than 2 mirrored copies
• CPC Class Codes: G06F 11/2064 while ensuring consistency
• CPC Class Codes: G06F 11/2087 with a common controller

Abstract

An additional communications link between two mass-storage devices containing LUNs of a mirrored-LUN pair, as well as incorporation of a fail-safe mass-storage-device-implemented retry protocol to facilitate non-drastic recovery from communications link failures within the controllers of the two mass-storage devices, prevents build-up of WRITE requests in cache and subsequent data loss due to multiple communications-link and host computer failures. The combination of the additional link and the retry protocol together ameliorates a deficiency in current LUN-mirroring implementations that often leads to data loss and inconsistent and unrecoverable databases.

Claims

1. A reliable computing and data-storage system comprising:

a first host computer linked to a second host computer by a first communications medium;

a first mass-storage device, linked to the first host computer by a second communications medium;

a second mass-storage device, linked to the second host computer by a third communications medium;

a fourth communications medium directly linking the first mass-storage device to the second mass-storage device for transmission of sequenced data written to a dominant logical unit within the first mass-storage device to the second mass-storage device for writing to a remote-mirror logical unit;

a first link that directly links the first mass-storage device to the first communications medium;

a second link that directly links the second mass-storage device to the first communications medium;

controller logic within controllers of the first and second mass-storage devices, the controller logic within at least one controller detecting a failure of the fourth communications medium, the controller logic within both controllers cooperating to retransmit any lost data from the first mass-storage device to the second mass-storage device by way of the first and second links and the first communications medium.

2. The system of claim 1 wherein the first mass-storage device receives WRITE requests directed to the dominant logical unit, executes the WRITE requests with respect to the dominant logical unit, and places the WRITE requests in an output memory buffer for subsequent transmission to the second mass-storage device.

3. The system of claim 2 wherein the second mass-storage device receives WRITE requests transmitted from the first mass-storage device and stores the WRITE requests in an input memory buffer for subsequent execution with respect to the remote-mirror logical unit.

4. The system of claim 3 wherein, when the fourth communication medium fails, and WRITE requests cannot be transmitted by way of the first and second links and the first communications medium, the controller of the first mass-storage device purges WRITE requests from the output memory buffer, noting data units of the remote-mirror logical unit that would have been written by the purged WRITE requests in a bit map.

5. The system of claim 3 wherein, after the controller of the second mass-storage device detects a failure to receive a sequenced WRITE request from the first mass-storage device and requests retransmission, from the first mass-storage device, of the WRITE request that was not received by the second mass-storage device, and the WRITE request is still not received after a threshold period of time, the controller of the second mass-storage device purges WRITE requests from the input memory buffer, noting data units of the remote-mirror logical unit that would have been written by the purged WRITE requests in a bit map.

6. A reliable computing and data-storage system comprising:

a first host computer linked to a second host computer by a first communications medium;

a first mass-storage device, linked to the first host computer by a second communications medium;

a second mass-storage device, linked to the second host computer by a third communications medium;

a fourth communications medium directly linking the first mass-storage device to the second mass-storage device for transmission of sequenced data written to a dominant logical unit within the first mass-storage device to the second mass-storage device for writing to a remote-mirror logical unit;

a first link that directly links the first mass-storage device to the first communications medium;

a second link that directly links the second mass-storage device to the first communications medium; and

controller logic within controllers of the first and second mass-storage devices that detect a failure of the fourth communications medium and transmit any lost, and all subsequent, data for writing to the remote-mirror logical unit from the first mass-storage device to the second mass-storage device by way of the first and second links and the first communications medium.

7. The system of claim 6 further including:

controller logic within a controller of the second mass-storage device that detects a failure to receive sequenced data from the first mass-storage device and requests retransmission, from the first mass-storage device, of the data that was not received by the second mass-storage device.

8. The system of claim 7 wherein the first mass-storage device receives WRITE requests directed to the dominant logical unit, executes the WRITE requests with respect to the dominant logical unit, and places the WRITE requests in an output memory buffer for subsequent transmission to the second mass-storage device.

9. The system of claim 8 wherein the second mass-storage device receives WRITE requests transmitted from the first mass-storage device and stores the WRITE requests in an input memory buffer for subsequent execution with respect to the remote-mirror logical unit.

10. The system of claim 9 wherein, when the fourth communication medium fails, and WRITE requests cannot be transmitted by way of the first and second links and the first communications medium, the controller of the first mass-storage device purges WRITE requests from the output memory buffer, noting data units of the remote-mirror logical unit that would have been written by the purged WRITE requests in a bit map.

11. The system of claim 9 wherein, after the controller of the second mass-storage device detects a failure to receive a sequenced WRITE request from the first mass-storage device and requests retransmission, from the first mass-storage device, of the WRITE request that was not received by the second mass-storage device, and the WRITE request is still not received after a threshold period of time, the controller of the second mass-storage device purges WRITE requests from the input memory buffer, noting data units of the remote-mirror logical unit that would have been written by the purged WRITE requests in a bit map.

12. A method for increasing the reliability of a computing and data-storage having a first host computer linked to a second host computer by a first communications medium, a first mass-storage device, linked to the first host computer by a second communications medium, a second mass-storage device, linked to the second host computer by a third communications medium, and a fourth communications medium directly linking the first mass-storage device to the second mass-storage device for transmission of sequenced data written to a dominant logical unit within the first mass-storage device to the second mass-storage device for writing to a remote-mirror logical unit, the method comprising:

providing a first link that directly links the first mass-storage device to the first communications medium;

providing a second link that directly links the second mass-storage device to the first communications medium;

detecting, by a controller within the second mass-storage device, a failure to receive sequenced data from the first mass-storage device by the second mass-storage device and requesting retransmission, by the controller within the second mass-storage device from the first mass-storage device, of the data that was not received by the second mass-storage device; and

detecting a failure of the fourth communications medium by a controller within the first mass-storage device and transmitting any lost, and all subsequent, data for writing to the remote-mirror logical unit from the first mass-storage device to the second mass-storage device by way of the first and second links and the first communications medium.