Browse Prior Art Database

Preservation of error information via staging buffer

IP.com Disclosure Number: IPCOM000013696D
Original Publication Date: 2001-Mar-01
Included in the Prior Art Database: 2003-Jun-18
Document File: 2 page(s) / 39K

Publishing Venue

IBM

Abstract

Disclosed is a system including means for preserving information via a staging buffer. In a storage subsystem, in the event of a failure, it is important to capture as much information as possible at the first occurence. The data to be recorded can run to many megabytes. If this data can be recorded at the first occurrence of the problem it alleviates the necessity to reproduce the problem and thus reduces customer down time. Current adapters solve this problem by writing this data to external data storage (disk drive) at the time of the problem. The adapter is then reinitialised and customer operations are resumed. However if the external data storage is inaccessible at the time when the data is to be stored it must be discarded and the information is lost. If the external data storage is unavailable at the time of the problem, this invention ensures that the dump data persists until the external data storage becomes available.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Preservation of error information via staging buffer

Disclosed is a system including means for preserving information via a staging buffer. In a storage subsystem, in the event of a failure, it is important to capture as much information as possible at the first occurence.

     The data to be recorded can run to many megabytes. If this data can be recorded at the first occurrence of the problem it alleviates the necessity to reproduce the problem and thus reduces customer down time.

     Current adapters solve this problem by writing this data to external data storage (disk drive) at the time of the problem. The adapter is then reinitialised and customer operations are resumed. However if the external data storage is inaccessible at the time when the data is to be stored it must be discarded and the information is lost.

     If the external data storage is unavailable at the time of the problem, this invention ensures that the dump data persists until the external data storage becomes available.

     During normal operation the adapter stores customer data in a data buffer. Part of the buffer is allocated to write cache data which must be retained across a reset or power outage. However a second part of the buffer is used for read cache data; this data is stored solely to improve performance of read commands and may safely be discarded at any time.

     According to this invention, when the adapter encounters an unexpected problem, a dump function on the adapter discards this read cache data and the buffer space is reused for storing the dump information. Following this the adapter is reinitialised; in most cases this reinitialisation will restore accessibility of the disk drives (since whatever problem caused the adapter to stop running will normally be cleared, the hardware and firmware will be restarted and the storage subsystem should be capable of operating normally). Since the disk drives can now be accessed once more, the dump information can be copied from the adapter buffer to a disk drive. The dump information in the buffer is then discarded and the buffer space is once more made available for read caching. Consequently, subsystem performance quickly returns to its normal level. Using this invention the dump information is retained during the reinitialisation and is present on a disk drive after the subsystem restarts, even though the adapter may have been unable to access the disk dri...