STATISTICAL AND ENVIRONMENTAL DATA the result that the pack is unable to be read from the
LOGGING SYSTEM FOR DATA PROCESSING disk drive on which it is currently mounted. Therefore,
STORAGE SUBSYSTEM it is sometimes impossible to distinguish whether errors
„ - . .I1TI^ „.... ^^n., in data transfer to or from a given drive are due to the
BACKGROUNDOFTHEINVENTION 5 being m error or to the disk pack being in error.
In modern day computer systems a central „ „
processing unit, or CPU, processes instructions and SUMMARY OF THE INVENTION
data, most of which, due to main storage limitations The present invention avoids the above shortcomings
within the CPU, are stored in one or more peripheral by providing a statistical record of usage and error in
storage devices external to the CPU. Generally, a CPU 10 formation for each physical device in a subsystem and
is connected to a data channel which, in turn, is con- for each physical volume on the physical device,
nected to the peripheral storage devices by way of a Briefly, the invention provides counters for counting
storage control unit. An operation performed at the the number of bytes of data read and the number of ac
CPU or channel is said to be performed at the system cess motions, for each physical device and correlates
level, while an operation performed at the peripheral these to the number of correctable data errors, uncor
storage device or storage control unit is said to be per- rectable errors, and access motion (or seek) errors for
formed at the subsystem level. a gjven physical volume within the physical device.
A request for transfer of data between a peripheral when ^ number of errors of at least one type exceeds storage device and the CPU is generally in the form of a 2Q a ^,^4 number & compared to usage of at least
command stored in CPU main storage, the command one ^ ^ information is offloaded to
being termed a channel command word (CCW). A plu- ^ system by physical drjve ID wd Vo,ume ,D ^
rality of such requests in sequence are termed a chain b associating error information to volume and physi
of CCWs which result in a plurality of operations such ^ ^ it ^ iWe tQ infer ^ m enof occurri in as data transfers between the peripheral storage device 25 ^ subs tem is more ,ikel in tne n ical vo,ume or
and the CPU. In the past, whenever an error wasen- fa a h^ ... UR this information is of
countered dunng data transfer from a chain of CCW s, ... tf g ... ^ m ^
the storage control unit would signal a data check com- exceeding its threshold,
munication to the channel, resulting in an interrupt to „_ .. . __
■ , ... ... ■ . • Whenever offloading occurs due to error overflow,
the CPU with the result that the entire chain of CCW s 30 . . ., , • , u . j
tJ, . . . . • ■ t_ * detailed diagnostic information is collected the next ar
would be re-executed from the beginning, in hopes of , . 7; . . . . . ,
... . . . <■ ... . „ .. . bitrary number of times an error of the type causing the
achieving data transfer without error. Recently, 1m- j- • . . j t. - ^ ■ • .
. . ■ _ j . »u ._ J offloading is encountered, and such information is used provements have been made to the system under
j. . . . for diagnostic purposes.
discussion, wherein when an error occurs in an opera- z-^w w: J J . J .
... , ■ - r ~r~,„ .. . „, Other objects and attendant advantages of this invention resulting from the chain of CCW s, the storage 35 .. .„ .J ... . & .
control unit has the ability to retry that particular CCW "on w11 become appreciated as the same becomes
without re-executing the entire chain of CCW's and in ***** understood by reference to the following
such a manner that the retry of the CCW appears to the d^lf descnption when considered in connection
system merely as a normal CCW fetch, as opposed to w.th the accompanying drawmg
being a system interrupt. While this improvement has 40 *}G [ls[au representation of a storage subsystem
had the effect of significantly improving system within which the invention can be embodied.
throughput and efficiency, it has raised a problem in n0 2 IS a representation of various parts of a data
that now the system has no way of knowing the en- stora«e svstem shows the manner in which the in"
vironmental status and statistical error and usage status ven5°n to embodied thereir
of the peripheral storage devices, inasmuch as most er- 45 TM3. 3 is a representation of the error and usage
rors are handled at the subsystem level, without system counters of the invention.
intervention. FIG. *ls a representation of the manner in which the In the system of the type under discussion, the counters of FIG. 3 may be laid out in the writeable conperipheral storage devices are generally of the type trol storage in the storage control unit of the subsystem, having a removable storage medium termed a volume. 5<> 17,0 5 f a representation of the manner in which the For example, the peripheral storage devices may be system is informed that an intolerable number of errors rotating disk storage drives which have removable disk has occurred for a given physical volume, packs as the storage volumes; or they may be tape F105- 6A and 6B are flowcharts illustrating the drives which have removable tapes as the storage method of our invention.
volumes; or other like devices. This being the case, and 55 FIG. 7 is an illustration of a summary record useful in
taking rotating disk storage drives as an example, a disk our invention.
pack may be written on a first drive and read from a Before beginning a description of the invention, it second drive. Disk packs may be therefore in- would first be well for background purposes to review terchanged from one drive to another to yet another. information storage generally in one system in which When an inordinate number of errors occur during a the current invention may find use, it being recognized data transfer or other type operation to or from a given that the invention will also find use in other types of drive, the drive may become suspect as being in error. storage systems. Information is generally stored, in the However, it is possible the error may actually be in the system under discussion, on disk pack volumes on medium, i.e., in the disk pack itself. That is, the record- 6J tracks, in records comprising three information fields: a ing medium may have been damaged; or perhaps the count field, a key field, and a data field. The beginning pack was written on another disk drive which may have of a record is indicated, for control purposes, by an adbeen out of tolerance through wear, for example, with dress marker. Each address marker is preceded by a