1 karl 1.1.2.1 Using the CIM/XML Pull Operations
2
3 STATUS
4
|
5 karl 1.1.2.4 <<< The TODO section is being maintained during the review and checkin process
|
6 karl 1.1.2.1 to keep track of problems, errors, notes, etc. Must be deleted before
7 checkin to head of tree. Please feel free to add notes, etc in this
8 section as you review/test.>>>>>>
9
|
10 karl 1.1.2.4 TODO list:
|
11 karl 1.1.2.3 1. Binary operation from OOP. Need to add counter to binary
12 protocol to be able to count objects in response. Generates
13 warnings in things like messageserializer and does not work with
|
14 karl 1.1.2.31 OOP right now. Fixed by converting to XML. Concluded that we do not
15 need to do this. The binary response is not really used often
16 in the current environment So double mapping it is not a major issue.
17 Leave this as FUTURE
|
18 karl 1.1.2.23 2. OpenExecQuery - Code is incomplete in that it does not include the
|
19 karl 1.1.2.31 return from the exec query function to the aggregator yet. Code for Pull
20 part of OpenQueryInstancesRequest a) should be part of
|
21 karl 1.1.2.4 the common CIMOperationRequestDispatcher execCommon code.
|
22 karl 1.1.2.31 3. The changes to WQLCIMOperationRequestDispatcher and CQL... for handling
|
23 karl 1.1.2.4 pull not completed so we feed the responses back to the EnmerationContext
24 queues
|
25 karl 1.1.2.31 4. Minor TODOs, diagnostics, etc. still in the code
26 5. Extension to avoid double move of objects in CIMResponseData (one
|
27 karl 1.1.2.3 into enumerationContext queue and second to new cimResponseData for
28 response. Want to avoid second move by extending Open/Pull response
29 messages to include count and CIMResponse data to count objects out
|
30 karl 1.1.2.23 of queue when converting (avoids the second move). This would mean
31 extending the output writers to allow the count field to be supplied
|
32 karl 1.1.2.30 so they would only create output for up to the count supplied.(Make
33 this future beyond bug 9676)
|
34 karl 1.1.2.31 6. Add more static tests (currently only OpenEnumerateInstances and
|
35 karl 1.1.2.30 OpenEnumerateInstanceNames covered).
|
36 karl 1.1.2.26 Low priority because most of this except exact XML covered in other
|
37 karl 1.1.2.28 tests. note that we can only add very limited tests since a) no
|
38 karl 1.1.2.31 way to test multiple operations in sequence(i.e. open/pull) and
|
39 karl 1.1.2.28 cannot be sure how many responses will come back in all cases even
40 for open.
|
41 karl 1.1.2.32 7. Clean up the CIMOperationRequestDecoder.cpp rejectCIMException
|
42 karl 1.1.2.31 further since there is some overlap. VERY Low priority.
|
43 karl 1.1.2.32 8. Correct issue between operations and HTTP where we are sending
|
44 karl 1.1.2.31 trailers with exceptions. Modify response mechanisms so that we
45 set non-chunked for all responses where we send error responses to
46 avoid the trailers.
|
47 karl 1.1.2.32 9. It would be more efficient in dispatcher to always use exception for
|
48 karl 1.1.2.31 rejects and change the _reject functions so that they never return
49 when they reject. This is VERY LOW PRIORITY and primarily saves
50 a few lines of code in the reject functions and their calls. Means we
51 would code.
52 _rejectIfEnumerationToBroad(...)
53 _rejectIfThisParameterIncorrect(...)
54
55 instead of
56 if (_rejectIfEnum...)
57 {
58 return true
59 }
60 It would mean that the method trace for the handlers would not return
61 an exit if we rejected. LOW PRIORITY
|
62 karl 1.1.2.32 10. There are still a couple of template functionsaround the task of
63 distributing requests to the multiple providers.
64 In fact there are two similiar but different templates for
|
65 karl 1.1.2.31 the associators(i.e. assoc, ref, etc.) functions and the parallel
66 openAssoc functions. It would be nice to consolidate that logic and
67 further to try to create a non-template form for those functions. LOW
68 PRIORITY.
|
69 karl 1.1.2.32 11. External runtime variables. Proposing that they be fixed for this release
|
70 karl 1.1.2.31 rather than set by configuration. Making this a separate bug.
71 See bug 9819 for the changes to cover this.
|
72 karl 1.1.2.26
73 NEXT TASKS:
|
74 karl 1.1.2.32 a. finish 8, 11 3 above
75
76 26 May 2014
77 1. Remove a number of diagnostics.
78 2. Make cache size in EnumerationContextTable dynamic and remove function
79 to set default values. MaximumEnums set from Dispatcher on constructor for
80 EnumerationContextTable and table size set to 1/2 of maximum number of
81 open contexts.
82 3. Clean up some statistics in EnumerationContextTable
83
|
84 karl 1.1.2.31
85 9 May 2104
86 1. Converted several template functions in CIMOperationRequestDispatcher
87 to real functions
88 2. Added intermediate CIMMessage for CIMPullOperationRequestMessage since
89 all of the pulls have exactly the same data. This allowed us to
90 convert the template function processPullRequest from template to
91 real function
92 3. Added intermediate CIMMessage for CIMOpenOperationRequestMessage betweeh\n
93 all of the Opens and the CIMRequestData message. This will allow
94 common code to test all of these variables in all 6
95 CIMOperationRequestDispatcher handleOpen... functions
96 3. Added variable in CIMOperationRequestDispatcher handleEnqueue to make
97 releasing the original request conditional on return from the handle...
98 for open... and pull... handlers.
99 This allowed us to drop one place where we were creating a new request msg
100 for the delayed response and use the original request. This meant that
101 we could convert the issueOpenOrPullResponse from template function to
102 real function.
103 4. Moved open and pull constructors in CIMMessage.h to CIMMessage.cpp. NOTE:
104 openQueryInstances not moved. Will do when we get this funciton running.
105 karl 1.1.2.31 5. Added test for correct namespace in pull operations. Note that it is
106 never really used but since supplied with request, we test to be sure it is
107 same as open request namespace. added internationalized exception for this.
108 6. We should probably add same test to close just as an input parameter
109 validation.
110 6. Modified processPullRequest to close the enumeration context if
111 incorrect response type received. Before we just returned exception and
112 left context open for next request. Closing is much more logical since the
113 client really did send an erronous request.
114 7. Modified PullErrors to correctly test above changed in behavior.
115 8. Removed a significant number of diagnostic messages.
116
|
117 karl 1.1.2.23
|
118 karl 1.1.2.30 4 May 2014
119 1. Clean interface with EnumerationContext timeout thread
120 2. Remove more diagnostics
121 3. Correct issue with enumerationContext Release that was causing
122 failures during nightly tests.
123
|
124 karl 1.1.2.29 2 May 2014
125 1. Remove Diagnostics from ResponseData.cpp
126 2. Correct setup and start of EnumerationContextTimerThread
127 3. Change limit on max number of enum contexts to be set from Dispatcher.
128 4. Increase size of hashtable for enum contexts.
129 5. Corrected issue causing failures in removeContext. Now runs through nightly
130 tests most of the time.
|
131 karl 1.1.2.28 1 May 2014
132 1. Cleanup of code and variable names.
133 2. Removed namespace variable from EnumerationContext. It was never used.
134 3. Added test for max number of open enumerations. Right now it is fixed
135 and set about 100. Note that this is tied to the size of hash table
136 so that there is a issue with more dynamic setting of max and the
137 hash table size.
138 4. Removed several unused functions.
139 5. Found one possible issue causing thread conflict but not the core one
140 yet.
141
142 30 April
143 1. Added exception response to dispatcher for Limits exceeded on
144 CreateEnumerationContext. Returns Exceeded server limits CIMException.
145 2. Fixed possible deadlock where we were not unlocking EnumerationContext
146 under error condition
147
148
|
149 karl 1.1.2.27 27 April - mergeout to head of tree
150 1. Mergeout of the code to the head of tree. Documented that mergeout in
151 the bug
152
153 27 April 2014
154 1. Clean up in a number of areas, mostly just code without changing
155 functionality.
156 2. Modified CIMMessage.h slightly to clarify names of the pull intermediate
157 messages (the common open and pull response);
158 3. Removed a number of diagnostics from CIMResponseData
159 4. Added new function to ResponseStressTestProvider to be able to terminate
160 output after a defined number of objects with a CIMException.
161 5. Removed a number of other TODOs
162 6. Corrected a couple of errors in the server message bundle.
163 7. Modified ResponseEncoder to only output the error for pull errors.
164 It was errounously outputting the EndOfSequence, EnumerationContext and
165 the error. This way, only the error is sent which is spec.
166 8. Removed number of TODOs and cleaned up a couple of others.
167 9. Added static tests for OpenEnumerateInstanceNames.
168 10. Modified a number of the temporary diagnostics to be permanent by
169 changing them to PEGASUS_DEBUG_ASSERTs
170 karl 1.1.2.27
|
171 karl 1.1.2.24 22 April 2014
172 1. Clean up some messages.
173 2. Still trying to find issue that randomly keeps messages in timeout
174 queue long after they have been closed and returned.
|
175 karl 1.1.2.25 3. Fixed memory loss error.
|
176 karl 1.1.2.26 4. Added zero reject for OperationTimeout because that effectively disables
177 the whole timeout mechanism so that enumerations could be started, never
178 completed by the client and would just hang around in memory. That
179 would be a significant DDOS attack (start very large enum with no
180 timeout) and never do pulls after open.
|
181 karl 1.1.2.25
|
182 karl 1.1.2.23 21 April 2014
|
183 karl 1.1.2.27 1. Reorganized code in CIMOperationRequestDecoder.cpp ...Iparam classes.
|
184 karl 1.1.2.23 Cleaned up the rejectCIMException code.
185 2. Cleanup some issues in the EnumerationContextTimeout Thread.
186 3. Fixed issue in pullop where it was using zero as default for default
187 operation timeout where it should have been NULL, the DMTF specified
188 default.
189 4. Enabled the test to reject 0 as operation timeout on a request and but
190 controled by a #define in CIMOperationRequestDispatcher.cpp
191 5. Added test for excessive new OperationContext entries in table.
192
193 18 April 2014
194 1. Internationalize remaining messages in CIMOperationRequestDecoder.cpp and
195 clean up the decoders for some of the messages so they use the
196 ...Iparam classes. Note that we did not clean up all of the old message
197 decoders, just the new ones and we did not clean up the Count message
198 since it is deprecated. TODO clean these up so the decoder is really
199 consistent.
200 2. Enabled the EnumerationContext Timer and removed code for alternate
201 non-thread solution.
|
202 karl 1.1.2.16
|
203 karl 1.1.2.21 17 April 2014
204 1. Reorganized the lock/unlock for handling the EnumerationContext because
205 there were thread windows (i.e. using the context after it was released
206 by other thread).
207 2. Modified code for the InteroperationTimerThread and deleted old code that
208 allowed this whole thread to by bypassed.
209 3. Internationalized a number of CIMOperationRequestDispatcher CIMException
210 response messages.
|
211 karl 1.1.2.22 4. Fixed stress tests in Client/tests/pullop that were getting timeout. The
212 issue was not the pull code but that fact that the idle client timeout
213 appears to be measureing to the end of the operation and the operation
214 was taking more than 40 seconds which is the default timeout. The idle
215 timeout takes no account of the fact that chunks are being regularly
216 delivered.
217 5. Several minor code changes.
218 NOTE: At this point the code passes extensive tests of the pull operations
219 both in OOP and none OOP mode (but with lots of diagnostics that may
220 be affecting timing).
|
221 karl 1.1.2.21
|
222 karl 1.1.2.20 10 April 2014
223 1. Fix issues causing failure with repeated pull operations.
224 2. Consolidate some of the CIMOperationRequestDispatcher.cpp template code
225
|
226 karl 1.1.2.19 7 April 2104
227 MergeOut of head of tree.
228
229 7 April 2014
230 1. Permanently fixed issue with getting correct number of objects on response.
231 This was introduced when did a temp fix to the OOP processing, in particular
232 removed the condition variable wait in the CIMOperationDispatcher handle
233 functions for opens and pulls. This moved the functionality to kick off
234 open and pull responses to the provider threads if there are not enough
235 objects when the handle is processed. Also added a variable to allow us to
236 test with either a) responses are required to satisfy the original
237 request size or) b, responses are required only to return some objects.
238 We will add statistics to see which of these works best.
239 2. Modified WsmProcessor EnumerationContext class since the name conflicts
240 with the pull Operation EnumerationContext class.
241 3. Created a new intermediate level of CIMMessage, the CIMPullResponseData
242 Message so that all of the open and Pull responses can use common code
243 since there are only two variables (endOfSequence and EnumerationContext)
244 and they are common across all the open and pull responses.
245 4. Removed a number of diagnostics.
246 5. Cleaned up the Dispcatcher so that the open and pulls have common response
247 karl 1.1.2.19 code and that code can be used from the dispatcher handle functions and
248 the provider response functions.
249
|
250 karl 1.1.2.18 31 March 2014 - Checkin
251 1. Fixed issues in OOP processing of pull operations, in particular
252 issues with cimxml output format when processed through the
253 *InternalXmlEncoder functions.
254 2. Clean up some of the internalXml functionality
255 3. Found issues causing timeout with a particular provider. The issue
256 is that the dispatcher and monitor end up using the same thread so the
257 condition variable in the dispatcher thread stops the monitor. Turned
258 off the conditionVariable in getCache for the moment which means that
259 we get number of responses for open... with 0 objects before the
260 providers can begin to respond. This is only for test.
261 4. Added some statistics for enumerations and display the statistics
262 when we close the server (same as cache statistics)
263
264 12 March 2014 - Mergeout and Mergein
265 1. Mergeout to head of tree for this date and mergein for patch update
266 to bug 9676
267 2. Extensions to pullop tests program and tests.
268 3. Added some diagnostics in looking for OOP issue.
269 4. Removed a number of diagnostics messages and cleaned up code in
270 dispatcher to simplify pull operation processing.
271 karl 1.1.2.18
|
272 karl 1.1.2.17 15 December 2013
273 1. Mergeout and mergein up to 15 December 2013
274 2. Clean up issues from tests documented in bug 9676 last week.
275 3. Clean up some code in dispatcher
276 4. Remove the filter function from ResponseStressc++Provider.
277
|
278 karl 1.1.2.16 21 November 2013
279 1. Mergeout from head of tree to 21 November 2013.
|
280 karl 1.1.2.15
281 18 November 2013
282 1. Cleanup of a bunch of minor errors and completion of all of the code for
283 the openQueryInstances except for the PullInstances in Dispatcher and
284 the aggregator function.
285 2. OpenqueryInstances added to cimcli.
|
286 karl 1.1.2.14
287 13 October 2013 CVS branch update.
288 1. Integrated bug 9786 into the branch. Note that we need to test the
289 generated statistics.
290 2. Mergeout executed to update to head of tree as of 8:00 am 13 October 2013.
291 3. Cleaned up several errors in OOP processing. Note that there is at least
292 one issue left when we to a pull on ManagedElement in at least one of the
293 namespaces.
294 4. Cleaned up some of the outstanding diagnostic code
295 5. Generally passes all tests except for one test of pullop where it is trying
296 to pull enum instances CIM_ManagedElement from a particular namespace.
|
297 karl 1.1.2.12
|
298 karl 1.1.2.16 NOTE: I did not make comments here for changes in October despite the fact
299 that I did 2 mergouts, number of fixes, and a mergein.
300
|
301 karl 1.1.2.12 30 September 2013 - CVS Update
302 Mergeout head of tree up to 29 September 2013.
|
303 karl 1.1.2.10
304 29 September 2013. CVS update.
305 1. Modified calls to statisticalData.cpp to a) directly call with request
306 type, b) incorporate the open, pull, etc. messages. However, since these
307 are not part of the CIM class, we must do something special with them.
|
308 karl 1.1.2.13 See bug 9785 for full solution to this issue.
|
309 karl 1.1.2.10 2. Corrected OOP interface to enable new flag to indicate internal operations
310 and set host, etc.
311 3. Add code to CQLOperationsDispatcher and WQLOperationDispatcher to clean
312 up CIMResponseDataCounter after filtering.
313 4. Modified ProviderAgent to set Host info for some pull operations.
314 5. Added new flag to CIMBinMsgSerializer and Deserializer.
|
315 karl 1.1.2.8
|
316 karl 1.1.2.9 17 September 2013 CVS update (Actually two different updates over 3 days)
317 1. Clean up some issues in CIMMessage.h and CIMMessage.cpp
318 2. Extend OpenExecQuery to WQL and CQL processors but return not complete
319 3. Remove memory leak in EnumerationContext and EnumerationContextTable
320 handling.
321 4. Created template functions for much of the pull operations.
322 5. Reversed order of queryLanguage and query (and changed names to match
323 execQuery) in client and server. Note that these are the execQuery
324 WQL and CQL filters and NOT FQL filters.
325 6. Some code cleanup in dispatcher
326 7. Today, passes all tests in pullop but issue in alltests. For some reason
327 not finding CIMObjectManager instance. Also, leaves enumeration contexts
328 if client terminates since cleanup thread not operating.
|
329 karl 1.1.2.11 8. XML from OOP not correctly processed.
|
330 karl 1.1.2.9
|
331 karl 1.1.2.8 14 September 2013 CVS update
332 Merged out up to 25 August. Cleaned up all operations and standardized code.
333 At this point the non pull operations code is in a set of templates but the
334 pull is not yet.
335 Fixed a significant number of problems so that it appears that the operations
336 except for OpenExecQuery run stably, at least with the pullop test program.
337 Note that there is a problem in that the Interop control provider is not
338 returning its singleton wbemserver object for some reason. Causes a test
339 failure
|
340 karl 1.1.2.7
341 Fixed for 16 June CVS Update
342 1. Cleaned up the enumerationContext and Table release functions and tested
343 to confirm that we do not lose memory in either normal sequences or
344 sequences that close early. Cleaned up pullop and added more tests
|
345 karl 1.1.2.8 Taged Before: PREAUG25UPDATE and after POSTAUG25UPDATE
|
346 karl 1.1.2.4
347 Fixed for 9 June CVS update
348 1. Cleaned up code for OpenQueryInstances. Note that this is incomplete.
349 No support in WQL or CQL Operations
350 2.
351
352 What was fixed for 5 June checkin.
|
353 karl 1.1.2.3 1. Extended ResponseTest MOF for for both CMPI and C++ subclasses
354 2. Fixed issues with pullop.
355 3. Fixed temp issue with CIMResponseData size by putting in mutex. That
356 is not a permanent fix but it gets around issue probably in the control
357 of the move logic that meant counts were off.
358 4. Fixed issues in Dispatcher so that associator code works. Still messy
359 code in the dispatcher.
360 5. Changed name of Enumerationtable.h & cpp to EnumerationContextTable.*
361 6 Changed name of ResponseStressTest module, classes, etc.
362
363 TAG: TASK_PEP317_5JUNE_2013_2
364
|
365 karl 1.1.2.2 2 June 2013
|
366 karl 1.1.2.1
367 Issues - KS
368
|
369 karl 1.1.2.15 - Still way to many TODO and KS comments and KS_TEMPS. Removing bit by bit.
|
370 karl 1.1.2.1
|
371 karl 1.1.2.15 - Runtime variable connection for the config parameters not installed. That
372 has been made into a separate bug (see bug 9819)
|
373 karl 1.1.2.1
374 5. Issue with the threaded timer. For some reason during tests it
375 eventually calls the timer thread with trash for the parm (which is
376 pointer to the EnumerationTable object). Caught because we do a valid
377 test at beginning of the function.
378
|
379 karl 1.1.2.2 6. Still using the templates in CIMOperationRequestDispatcher to simplify
380 the handle... processing.
381
382 7. I think I have a way around the double move of objects in the
383 EnumerationContext so that the outputter will just take a defined number
384 of objects directly from the gathering cache and save the second move.
385
|
386 karl 1.1.2.15 8. Not yet passing all tests but getting closer now. The major test that is
387 causing an error today is the execution of a full enumeration with the
388 forceProviders = true. This causes a client timeout sometimes.
|
389 karl 1.1.2.2
390
391
|
392 karl 1.1.2.1 ===========================================
393
394 OVERVIEW:
395
396 The operation extensions for pull operations defined in the DMTF specification
397 DSP0200 V 1.4 were implemented in Pegasus effective Pegasus version 2.11
398 including Client and Server.
399
400 These operations extend the CIM/XML individual operations to operation
401 sequences where the server must maintain state between operations in a
402 sequence and the client must execute multiple operations to get the full
403 set of instances or instance paths.
404
405 The following new CIM/XML operations as defined in DSP0200 are included;
406
407 -OpenEnumerateInstances
408 -openEnumerateInstancePaths
409 -OpenReferenceInstances
410 -OpenReferenceInstancePaths
411 -OpenAssociatiorInstances
412 -OpenAssociatorInstancePaths
|
413 karl 1.1.2.16 -OpenQueryInstances
|
414 karl 1.1.2.1 -PullInstancesWithPath
415 -PullInstancePaths
|
416 karl 1.1.2.16 -PullInstances
|
417 karl 1.1.2.1 -CloseEnumeration
418 -EnumerationCount
|
419 karl 1.1.2.2 OpenExecQuery
|
420 karl 1.1.2.1
421 The following operations have not been implemented in this version of Pegasus:
422
423 -OpenQueryInstances
424
425 The following limitations on the implementation exist;
426
427 1. The filterQueryLanguage and filterQuery parameters are processed by
428 the Pegasus client but the server returns error if there is any data in
|
429 karl 1.1.2.2 either parameter. This work does not include the development of the
430 query language. Note that a separate effort to extend Pegasus to use
431 the DMTF FQL query language is in process.
|
432 karl 1.1.2.1
433 2. The input parameter continueOnError is processed correctly by the client
434 but the Pegasus server only provides for false since the server does not
435 include logic to continue processing responses after an error is
436 encountered.
437 This is consistent with the statement in the specification that use of
438 this functionality is optional and the fact that the DMTF agrees that all
439 of the issues of continuing after errors have not been clarified.
440
441 3. The operation enumerationCount is not processed by the server today since
442 a) really getting the count would be the same cost as the corresponding
443 enumeration, b) the server does not include a history or estimating
444 mechanism for this to date.
|
445 karl 1.1.2.2 NOTE: After a through review as part of the development of the next version
446 of CMPI we have concluded that this operation is probably not worth the
447 effort. Since it is optional, Pegasus will only return the unknown status
448 at this point
|
449 karl 1.1.2.1
450 Since the concept of sequences of operations linked together (open, pull, close)
451 is a major extension to the original CIM/XML operation concept of completely
452 independent operations several new pieces of functionality are implemented
453 to control interOperationTimeouts, counts of objects to be returned, etc.
454
455 TBD - Review this
456
457 CLIENT
458
459 The new operations follow the same pattern as the APIs for existing operations
460 in that:
461
462 1. All errors are handled as CIMException and Exception
463
464 2. The means of inputting parameters are the same except that there are
465 significantly more input parameters with the open operations and for the
466 first time operations return parameters as well as objects in the
467 response. Specifically the open and pull operations return values for
468 enumerationContext which is the identity for a pull sequence and
469 endOfSequence which is the marker the server sends in open and pull
470 karl 1.1.2.1 responses when it has no more objects to send.
471
472 The significant differences include:
473
474 1. Processing of parameters on responses (i.e. the endOfSequence and
475 enumerationContext parameters are returned for open and pull operations).
476
477 2. Numeric arguments (Uint32 and Uint64 include the option of NULL in some
478 cases so they are packaged inside classes Uint32Arg and Uint64Arg in the
479 client api.
480
481 3. The association and reference operations ONLY process instances. They do
482 not include the capability to return classes like reference and associator
483 do and therefore return CIMInstance rather than CIMObject.
484
485 4. Paths are returned in all cases (i.e OpenEnumerateInstances and
486 PullInstancesWithPath where they were not with EnumeratInstances.
487
488 5. The client must maintain state between operations in a sequence (using
489 the enumerationContext parameter).
490
491 karl 1.1.2.1 TBD- Are there more differences.
492
493
494 SERVER
495
496 The Pegasus server attempts to always deliver the requested number of objects
497 for any open or pull request (the specification allows for the server to
498 deliver less than the requested number of objects and specifically to return
499 zero objects on open). We felt that it was worth any extra cost in processing
500 to provide the client with exactly what it had requested.
501
502 The pegasus server always closes an enumeration sequence upon receipt of any
503 error from the providers, repository, etc. Therefore the server will reject
|
504 karl 1.1.2.2 any request that has continueOnError = true;
505
506 Expansion to allow the continue on error may be added in a future version.
507 In any case, the whole purpose of the continue on error is really to allow
508 input from good providers to be mixed with providers that return errors so
509 that generally this would mean simply changing the logic in the return mechanism
510 to not shutdown when an error is received from any given provider.
511
512 Generally we do not believe that the providers need to do much more in the
513 future to support the continueOnError other than possibly allowing the provider
514 to continue processing after it has received an error.
|
515 karl 1.1.2.1
516 PROVIDERS
517
518 This implementation requires NO changes to the existing providers. The
519 provider APIs operate just as they do with the original operations.
520
521 Because the server processing is different however, there may be some
522 behavior differences primarily because the client now controls the speed of
523 delivery of objects.
524
525 In previous versions of Pegasus, the server attempts to deliver objects as
526 rapidly as then can be put on the network. In the case of HTTP chunked requests
527 they are delivered in chunks of about 100 objects. The primary delay for the
528 providers was the processing of each segment through the server. The server
529 is blocked so that no other segment can proceed through the server until that
530 segment is processed and sent on the network.
531 In the case of non-chunkedresponses, they are completely gathered in the serve
532 and then delivered as one non-chunked response. There were no delays for the
533 providers, just lots of possible memory use in the server.
534
535 The responses from providers (delivered through the deliver(...) interface are
536 karl 1.1.2.1 gathered into segments of about 100 objects and this group of objects is moved
537 through the server to be delivered to the client.
538
539 However with the inclusion of the pull operations, The segments of objects
540 from the providers are cached in the server response path until the
541 maxObjectCount for that request (open or pull) and that number returned in a
542 non-chunked response. Thus, if the client is slow to issue pull requests,
543 the providers might be delayed at some point to reduce memory usage in the
544 server (the delay appears as slow response tothe deliver operation).
545
546 In other words, the time to process large sets of responses from the provider
547 now depends on the speed of handling the client.
548
549 It is important to remember in developing providers that the Pegasus server
550 can most efficiently process responses if they are passed from the provider
551 to the server individually or in small arrays of objects rather than the
552 provider gathering very large arrays of objects and sending them to the
553 server.
554
|
555 karl 1.1.2.2 NEXT GENERATION PROVIDERS
556 KS_TODO
557
|
558 karl 1.1.2.1 CONFIGURATION PARAMETERS
559
560 The server includes several configuration parameters to set limits on the
561 processing of pull operations. All of these configuration parameters are
562 compile time parameters rather than runtime.
563
564 1. Maximum value of minimum interoperation time. This parameter defines the
565 maximum time allowed between the return of an open or pull response and
566 the receipt of the next pull or a close operation before the server may
567 close the enumeration. The specification allows the server to set a
568 maximum interoperation time and refuse open requests that with requested
569 operationTimeout greater than that time.
570 CIM_ERR_INVALID_OPERATION_TIMEOUT
571
572 This value is set with the Pegasus environment variable
573 PEGASUS_PULL....
574
575 2. Maximum objects returned in a single open or pull operation. The server
576 can set a maximum limit on the number of objects that can be returned in
577 a single open or pull oepration with the maxObjectCount parameter.
578
579 karl 1.1.2.1 3. Whether the server allows 0 as an interoperation timeout value. The value
580 zero is s special value for the interoperationTimeout in that it tells the
581 server to not timeout any enumeration sequence.
582
583 With this value for interoperationTimeout, the only way to close an
584 enumeration sequence is to complete all of the pulls or issue the close.
585 If for some reason the sequence is not completed, that enumeration context
586 would remain open indefinitly. Since in Pegasus any open enumeration
587 context uses resources (the context object and any provider resposnes that
588 have not yet been issued in a response) it would appear that most
589 platforms would not want to allow the existence of enumeration contexts
590 that cannot be closed by the server.
591
592 4, maximum consecutive pull requests with 0 maxObjectCount. The use of the
593 pull operation with maxObjectCount set to zero could be used to keep an
594 enumeration context open indefinitly (this tells the server to restart the
595 interoperationTimeout but not send any objects in the response). Therefore the
596 specification allows for the server setting maximum limits on this behavior
597 and returning the error CIM_ERR_SERVER_LIMITS_EXCEEDED if this limit is
598 exceeded.
599 Note that this is maximum CONSECUTIVE pulls so that issuing a pull with
600 karl 1.1.2.1 a non-zero count resets this counter.
601
602 KS-TBD - Is this really logical since we can still block by just issuing
603 lots of zero request and an occansional request for one object.
604
605 Pegaus sets the value of this limit to 1000 and allows the implementer to
606 modify it with the PEGASUS_MAXIMUM_ZERO_OBJECTCOUNT environment variable.
607
608 5. Default operationTimeout -
609
610 The default of this parameter is to refuse operat
611
612 In the current release of Pegasus these are all compile time parameters.
|
613 karl 1.1.2.11
614
615 NOTES On working with task branch.
616
617 Merge out Process
618
619 To keep our TASK branch in sync with the current head of tree we need
620 to do a regular merge out. the TaskMakefile contains the makefile
621 procedures to do this efficiently. NOTE: Following these procedures is
622 important in that you are merging out new material each time you do
623 the merge out. If you were just to repeatedly merge out, you would be
624 merging previously merged changes a second time causing a real mess.
625
626 Start with new directory and put TaskMakefile above pegasus (needed so you
627 have this file for the initial operations.
628
629 make -f TaskMakefile branch_merge_out BNAME=PEP317-pullop ## takes a long time
630
631 This checks out current head, merges it into task branch and sets tags
632 for the mergeout. Note that at the end of this step this work is
633 part of the TASK... branch.
634 karl 1.1.2.11
635 NOW check for conflicts, errors, etc. that resulted from the merge.
636 Look for conflict flags, compare the results (I use linux merge as a
637 good graphic compare tool) and build and test. When you are satisfied
638 that the merge out is clean, you can commit the results to the TASK...
639 branch
640
641 To commit the work to this into Task branch
642
643 make -f mak/TaskMakefile branch_merge_out_commit BNAME=PEP317-pullop
644
645 or manually commit and finish as follows
646
647 cvs commit
648 make -f mak/TaskMakefile branch_merge_out_finish BNAME=PEP317-pullop
649
650 ## This last step is important since it cleans up temporary tags to prepare
651 you for the next checkout
652
653 COMPARE TASKBRANCH WITH HEAD
654
655 karl 1.1.2.11 In a new pegasus work space do same as above for merge out.
656
657 make -f TaskMakefile BNAME=PEP317-pullop
658
659 This produces a result which is all of the head merged into the branch.
660 A diff of this is all the new changes to the head of tree that you will
661 include into the merge.
662
|