CAN CSA-Z243 180B-1999 Information processing systems - Open Systems Interconnection - File Transfer Access and Management - Part 1 General introduction AMENDMENT 2 Overlapped ac.pdf

《CAN CSA-Z243 180B-1999 Information processing systems - Open Systems Interconnection - File Transfer Access and Management - Part 1 General introduction AMENDMENT 2 Overlapped ac.pdf》由会员分享，可在线阅读，更多相关《CAN CSA-Z243 180B-1999 Information processing systems - Open Systems Interconnection - File Transfer Access and Management - Part 1 General introduction AMENDMENT 2 Overlapped ac.pdf（15页珍藏版）》请在麦多课文档分享上搜索。

1、Amendment 2:1999 to National Standard of Canada CAN/CSA-Z243.180-89 Amendment 2: 1993 to International Standard IS0 8571-1:1988 has been adopted without modification as Amendment 21999 to CAN/CSA-Z243.180-89. This Amendment was reviewed by the CSA Technical Committee on Information Technology (TCIT)

2、 under the jurisdiction of the Strategic Steering Committee on Information Technology and deemed acceptable for use in Canada. July 1999 Information processing systems - Open Systems Interconnection - File Transfer, Access and Management - Part 1 : General introduction AMENDMENT 2 : Overlapped acces

3、s ( Reaf f i r med 2004) Systemes de traitement de /information - lnterconnexion de systemes ouverts - Transfert, acces et gestion de fichiers - Partie I : Introduction generale AMENDEMENT 2 : Chevauchement dacces Reference number IS0 8571-1 :1988/Amd.2:1993 (E) IS0 8571-1:1988/Amd.2:1993 (E) Forewo

4、rd IS0 (the International Organization for Standardization) and IEC (the International Electrotechnical Commission) form the specialized system for worldwide standardization. National bodies that are members of IS0 or IEC participate in the development of International Standards through technical co

5、mmittees established by the respective organization to deal with particular fields of technical activity. IS0 and IEC technical committees collaborate in fields of mutual interest. Other international organizations, governmental and non- governmental, in liaison with IS0 and IEC, also take part in t

6、he work. In the field of information technology, IS0 and IEC have established a joint technical committee, ISO/IEC JTC 1. Draft International Standards adopted by the joint technical committee are circulated to national bodies for voting. Publication as an International Standard requires approval by

7、 at least 75 % of the national bodies casting a vote. Amendment 2 to International Standard IS0 8571-1:1988 was prepared by Joint Technical Committee ISO/IEC JTC 1, Information technology. IS0 8571 consists of the following parts, under the general title Information processing systems - Open Systems

8、 Interconnection - File Transfer, Access and Management - Part 1 : General introduction - Part 2 : Virtual Filestore Definition - Part 3 : File Service Definition - Part 4 : File Protocol Specification - Part 5 : Protocol lrnplementation Conformance Statement Proforma 0 ISOAEC 1993 All rights reserv

9、ed. No part of this publication may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying and microfilm, without permission in writing from the publisher. ISOAEC Copyright Office Case postale 56 CH-1211 Genbve 20 Switzerland IS0 857 1-1 :I 988/Amd.2:

10、 1993 (E) Information processing systems - Open Systems Interconnection - File Transfer, Access and Management - Part 1 : General introduction AMENDMENT 2 : Overlapped access 0 Introduction Clause 0 provides an introduction to this amendment The text in this clause is not intended for inclusion in I

11、S0 8571 part 1- 0.1 General IS0 8571 part 1 defines in an abstract way the externally visible file transfer, access and management service. This amendment extends this service definition to incorporate the services offered by overlapped access. 09 Rationale The objective in introducing overlapped ac

12、cess is to allow more efficient access to structured files when a single initiator has a need to perform many reading and updating operations; the serial nature of the current RAM data transfer services introduces a significant control overhead if the FADUS are small. In this context, an FADU is sma

13、ll if its transmission time is comparable with the time to complete a confirmed service on the association (the associations round trip delay). other BDT adions. subject to restridions; this is called be performed in parallel, so that both diredions of data transfer are exploited at any one time. Re

14、quests are then taken from the queue whenever either direction of transfer becomes free. This is called concurrent access. co- . secondly,r8adandwriteadionscan The transfer of a single FAW, speufied in a single FREAD request has the same interpretation as in IS0 8571. The resultant effect on the vir

15、tual filestore of a set of overlapped requests using consecutive accBss shall be the same as that of the equivalent set of requests issued in series; the used then the resultant effect of a set of write actions on the virtual filestore, is also serializable. However, due to the nondeterminism introd

16、uced by the use of concurrent access, it is also possible that in some uses of the service, the data transferred as a result of a read action is not consistent with the current state of the file. service provided is serialii. If concurrent access is 1 Scope and field of application This amendment ma

17、kes no adiitions to clause 1. 2 References This amendment makes no for each phase a set of valid messages is defined in terms of state transitions or LOTOS events. If overlapped access is allowed then (during the open regime) there may be more than one phase valid at any time i.e. there may be a num

18、ber of protocol exchanges in progress. Add the following definitions and re-numbsr acwrdng at any instant during the data transfer there is a preferred direction of data flow for that bulk data transfer operation. 8.3 External file service and internal file service Replace second sentence of a): The

19、 transfer of file data is modelled in the external file service as being the result of error-free operations. AW note: Note - Pending the specification of presentation symmetric synchronisation services, recovery mechanisms will not be available for use during overlapped access. 8A Service classes a

20、nd functional units Repiace d): d) the access class, which allows the initiating entity to perform operations on the file access data units, providing for the manipulation of remote data; 9 Functions associated with the file service Replace second sentence of first paragraph: These mechanisms are de

21、signed to provide a way for a user to perform a co-ordinated set of actions without interference from aaxsses on concurrent assochtions. Rephce fi w of third paragr: Concurrency control is provided sdely to control the corred parallel execution of tasks across multiple associations. Because the cov

22、controls provided are with reSpedtorunaccaconconcurrentassocbtron * s,theused the concurrency control mechanisms does not affed the abilii of the user to overiap data transfer adions on individual associations. If the use of overlapped acoess is allowed during an open regime and FADU locking has bee

23、n enabled, then the responder overrides the present concurrency controls and sets them as required by the FADU lock (as long as they do not conflict with the locks set by other users). Data transfer actions that are requested by the same user over concurrent assouat ions will be subject to concumncy

24、 controls. 10 Service providers supporting FTAM Append to third pagraph: The support for recovery mechanisms during overlapped access requires the use of the Session symmetric synchronisatiin service. Note - Pending the SpedRCation of pmenwon . symmetric * mdlinotbeavailable 6- recovery- forusedurin

25、goverlapped -. If the symmetric synchronisation sewices of the Session Layer are not used, then the specifwition of the Session Layer also requires the FTAM protocol implementation to keep track of a session token controlling which entity is able to issue synchronisation points. 3 IS0 857 1 -1:1988/

26、Amd.2:1993 (E) Section two: Virtual Filestore - General Concepts 11 Virtual filestore 13 Constralnt sets This amendment makes no adlitions to clause 11. This amedmmtmdras IK) additions to clause 13. 12 File structures This amendment makes no additions to clause 12. 14 Document types This amen these

27、may be performed sequentially or be overlapped. 15.5 File open phase Add the fobwing dyrse and renumber: Replace hst sentence: 15.7 Ouoriapgod - * areprwidedtoaNowarebxationofihe It establishes the data transfer facilities, including any special presentation contexts required for the transfer, and T

28、w the bulk data transfer This to allow more effii accass to struc!wed files when an the us8 of overlapped access. hitiator has a need to perform my reading and updating operations; the serial nature of the FTAM data transfer services introduces a significant control overtwad if the FADUS are small.

29、In this context, an FADU is small if its transmission time k comparabie with the time to complete a confirmed service on the association (the associations 15.6 Data access phase Replace first sentence of second paragraph: mAA6 regime F Fi - F- file selection regime file open regime data transfer reg

30、ime In EAD-ATTRIBUTE F-CHANGE -ATTRIBUTE F SEECT F-CREATE store management F-DESELECT F-DELETE NITIALlZE F -TERMINA F-U-ABORT F-P-ABORT Figure 10 - File service regimes and related primitives 5 IS0 8571-1 :1988/Amd.2:1993 (El round trip delay). The overlapped access takes place within a constant set

31、 of presentation contexts established when the file is opened, or previously. Two degrees of overlapped access are defined: 1) consecutive accss provides the initiator with the ability to request data transfer operations whilst a previously requested data transfer regime is in progress. Thus, the in

32、itiator is able to generate a queue of data transfer requests. ALSO, subject to the constraints defined in IS0 8571-3, the transfer of data for one data transfer request may be overlapped with the final exchange of PCI for the preceding data transfer operation. The resultant effect on the virtual fi

33、lestore of a set of overlapped requests using consecutive access is the same as that of the equivalent set of requests issued in series; the service provided is serializable. 2) concurrent access provides an initiator with the ability to overlap read data transfers with write data transfers. A serie

34、s of read data transfers may be overlapped consecutively as specified in 1); a series of write data transfers may be overlapped consecuthrely as specified In 1). Thus, the initiator is able to generate separate queues of read and write data transfer requests which are processed independently. That i

35、s, the PCI relating to one read data transfer may be overhpped with the exchange of PCI or file data for another read data transfer, similarly for write data transfers. This is subject to the constraints defined in 1so 8571-3. The msubntgffed of a set of write actions on the virtual filestore. is oe

36、rializable . However, due to the non- determinism introduced by the use of concurrent access ,it is possible that, in some uses of the service, the data transferred as a msuk of a read action is not consistent with the current sta6e of the file. 16 Mechanisms in the file protocol 16.4 Checkpoint ins

37、ertion Add to end of first pagraph: If overlapped access is allowed then the checkpointing mechanisms are independent for the two directions of flow, but not for data transfers in the same direction. 6 IS0 8571-1:1988/Amd.2:1993 (E) Annex A Examples of the use of FTAM (This annex does not form part

38、of the sfandard.) Replace clause 3: A3 Remote database access A3.1 Without overlapped access Suppose a user wishes to make a series of enquiries from a remote dabbase systom Tho steps taken are as follows (see figwe 12): a) initialise the associatin as in steps (a) and (b) of A.2, but selecting the

39、access cbss with the read and grouping functional units. b) perform the file initialisation steps as in (c) and (d) of A.2. Instead of providing a speclflcation of a user file, the initiating system names a body of information within the database addressed. c) the initiating system sends a request t

40、o read data, using the database query as the identifier of a file access data unit in a notional flat file of possible responses. d) the responding system returns the result of the query as if it were the contents of the requested file access data unit. e) the initiating system indicates the end of

41、the transaction and the responding system confirms it. f) further queries are processed similarly as necessary. g) the initiating system releases the linkage with the information base as in steps (9) and (h) of A2. h) the association and its supporting resources are released as in steps (i) to (k) o

42、f A2. Note -The above changes represent a defect in the base standard. They have been induded here krcompleteness. 1132 With consecutive access Suppose a user wishes to make a number of independent enquiries from a remote database system. The use of consecutive access can provide the user with a mor

43、e efficient service. The steps taken are as follows (see figure 13): a) initialise the assocktion as in steps (a) and (b) of A.2, but selecting the access class with the read, grouping and consecutive access functional units. b) perform the file initialisation steps as in (c) and (d) of A.2. Instead

44、 of providing a specification of a user file, the initiating system names a body of information within the database addressed. c) the initiating system sends a series of requests to read data. In each case the database query is used as the identifier of a file access data unit in a notional flat fil

45、e of possible responses. d) the responding system returns the result of each query as if it were the contents of the requested file a#;8ss data unit; the transmission of each result may begin after the end of data transfer marker has been sent for the previous request e) upon receipt of the relevant

46、 end of data transfer the initiating system indicates the end of the transaction; this is confirmed by the responding system. f) if a query is dependant on the information received as a result of a previous query, the initiating system need only wait until the receipt of the end of data transfer mar

47、ker for the previous query, before issuing the next request. g) further queries are processed similarly as necessary. 7 IS0 8571 -1 :1988/Amd.2:1993 (E) F-READ quest (1) F-READ quest (2) F-DATA indication (1) F-DATA indication (1 ) F-DATA indication (1) FDATA-END indication (1) F-DATA indication (2)

48、 F-DATA indication (2) F-DATA indication (2) F-TRANSFER-END quest (Rl) F-TRANSFER-END confirm (Rl) FDATA-END indication (2) F-TRANSFER-END request (R2) F-TRANSFER-END confirm (R2) F-READ indication(1) F-DATA qusSt (1) F-READ indication (2) F-DATA vuest (1) F-DATA 4ue5t (1 ) F-DATA-END request (1) F-

49、DATA request (2) F-DATA quest (2) F-DATA 9-t (2) F-TRANSFER-END indication (Rl) F-TRANSFER-END response (Rl) F-DATA-END request (2) F-TRANSFER-END indication(R2) F-TRANSFER-END response (R2) Figure 13 - Remote database access with consecutive access 8 IS0 8571-1:1988/Amd.2:1993 (E) h) the initiating system releases the linkage with the information base as in steps (g) and (h) of A2. I) the associatiin and its supporting resources are released as in steps (i) to (k) of A2. Add mure 13 and re-number accordhgly: A4 A local area network fileserver Rephce first paragraph: AA.1 Wahout overl