Custom HDB parametrizations not working as expected when in distributed project

[a.decelis]

Hello,

Using WinCCOA 3.14 P000 and Windows7.

I am working in this project where we want to run our WCCOAvalarch managers with custom numbers.
Succesfully tested the customizations in a non-distributed but redundant project. But applying same changes to distributed-redundante project, does not work.

TESTING
Testing machine (redundant project), ignoring help's advice, I use a script to pre-create the datapoints with the desired arNumbers (201 to 205):

And after creating the DPs, used the panels to configure the archives.

I add the corresponding WCOOAvalarch managers, with "-num"s from 201 to 205, started them. And modified some random DPEs' _archiv config to save values to on of the new archives (number 201).
Values are archived and SELECTs can be performed, everything working as expected. Restart the project, and still everything fine. So far, so good. But...

DISTRIBUTED PROJECT
Repeat same steps in the actual project (distributed and redundant). Everything seems fine but, upon project restart, archiving is stopped: no values are stored. :ohmy:
Ok, lets go to PARA, select any of the _archive configured dpes: everything looks as it is supposed to be. Just click the "apply" button and archiving resumes for that very DPE. Got to apply changes for every DPE for it to be archived. :S
Restart the project and archiving is lost again.

Another variant: programatically create _ValueArchive DPs from 6 to 200 (empty), and then use the WinCCOA panel to create those from 201 to 205.

What can possibly be happening? Why is that same configuration works in a non-distributed project but does not in a distributed one?

Attaching a ZIP with screen caps and the panel with the code I used to create the _ValueArchive DPs. https://www.winccoa.com/fileadmin/image ... n_caps.zip

[leoknipp]

Pleae note that in a redundant system always the datapoints _ValueArchive_ and _ValueArchive__2 need to be created.
In your panel I cannot see that the dp-element .state is set. This dp-element defines if the archive manager can be started or not.

Do you get error in the log viewer when configuring the archive configs or when starting the project?
Are you maybe somewhere using system names in your scripts which do not fit the system name of the distributed system?

Best Regards
Leopold Knipp
Senior Support Specialist

[Gertjan van Schijndel]

Were the value archives running when you added the archive configs?

Since 3.14 P009 the adding/activating of archive configs during an offline archive have been improved.

[a.decelis]

Hi there,

been busy lately, unable to keep up with this issue.

Yes, I am aware that in redundant systems, some datapoints have a "_2" version. The problem occurrs also in a distributed non-redundant project. Thank you for pointing it out, though.

The panel did not set the .state DPE. Now it does, and the result is the same: it simply does not work. :S

I get several messages in the log, better see attachment, though no obvious error.

Regarding system names in scripts: these issues are easily reproducible in an empty test project without need of scripts.

My steps for the test:
1 Create fresh distributed non redundant project, and configure it for localhost.
2 Start the project and use valArchCreator to create archive datapoints for archives 201 to 205.
3 Configure archive 201 using the panels.
4 Add the archive manager for 201 and start it

Code: Select all

# STARTING THE WCCOAvalarch -num 201 for the first time

WCCILdata    (0), 2017.05.03 12:25:54.471, SYS,  INFO,        4, Connected to (SYS: 1 Data -num 203 CONN: 1) @ admin-THINK (::1)
WCCILdata    (0), 2017.05.03 12:25:54.473, PARAM,INFO,        1, Manager Start, HistOrgMain, connection from archive received - initializing, archive number is: 201
WCCILdata    (0), 2017.05.03 12:25:54.476, SYS,  INFO,        0, , Manager (SYS: 1 Data -num 203 CONN: 1) initialised
WCCOAvalarch (201), 2017.05.03 12:25:55.469, SYS,  INFO,        0, , Archive-directory was not present - will be created
WCCOAvalarch (201), 2017.05.03 12:25:57.661, SYS,  INFO,        6, Initialization by Data Manager finished
WCCILdata    (0), 2017.05.03 12:25:57.676, SYS,  INFO,        0, , Sending initial values to archive
WCCOAvalarch (201), 2017.05.03 12:25:58.661, SYS,  INFO,        0, , Ready for operation.

[/size]
5 Create two ExampleDP_Float-type dps: test_archive_1 and test_archive_201.
6 Add _archive config to both DPs, and configure accordingly.
7 Check usage statistics for Archive 201, everything OK.
8 Restart the project.

Code: Select all

WCCILdata    (0), 2017.05.03 12:35:35.445, SYS,  INFO,        4, Connected to (SYS: 1 Data -num 203 CONN: 1) @ admin-THINK (::1)
WCCILdata    (0), 2017.05.03 12:35:35.446, PARAM,INFO,        1, Manager Start, HistOrgMain, connection from archive received - initializing, archive number is: 201
WCCILdata    (0), 2017.05.03 12:35:35.454, SYS,  INFO,        0, , Manager (SYS: 1 Data -num 203 CONN: 1) initialised
WCCOAvalarch (201), 2017.05.03 12:35:38.522, SYS,  INFO,        6, Initialization by Data Manager finished
WCCILdata    (0), 2017.05.03 12:35:38.526, SYS,  INFO,        0, , Sending initial values to archive
WCCILpmon    (1), 2017.05.03 12:35:38.567, SYS,  INFO,        9/pmon, The project is up and running. Entering monitor mode
WCCOAui      (1), 2017.05.03 12:35:39.027, SYS,  WARNING,    54, Unexpected state, Qt, QWindowsWindow::setGeometryDp: Unable to set geometry 1920x1058+-1920+-282 on QWidgetWindow/'QMainWindowClassWindow'. Resulting geometry:  1920x1054+-1920+-282 (frame: 8, 30, 8, 8, custom margin: 0, 0, 0, 0, minimum size: 519x457, maximum size: 16777215x16777215).
WCCOAvalarch (201), 2017.05.03 12:35:39.522, SYS,  INFO,        0, , Ready for operation.
WCCILdata    (0), 2017.05.03 12:36:35.905, SYS,  INFO,        0, , Loading DpIdentification
WCCILdata    (0), 2017.05.03 12:36:35.906, SYS,  INFO,        0, , (Re)Opening databases in multi user mode without transaction logging.
WCCILdata    (0), 2017.05.03 12:36:36.231, SYS,  INFO,        0, , DataManager, startBgMgr, (Re)starting WCCILdatabg
WCCILdatabg  (1), 2017.05.03 12:36:36.284, SYS,  INFO,        1, Manager Start, PROJ, test_distributed_project, V 3.14 - 3.14   platform Windows AMD64 linked at Feb  4 2016 21:46:19
WCCILdatabg  (1), 2017.05.03 12:36:36.285, SYS,  INFO,        3, Trying to connect to (SYS: 0 Data -num 0 CONN: 1) @ localhost:4897
WCCILdatabg  (1), 2017.05.03 12:36:36.286, SYS,  INFO,        4, Connected to (SYS: 0 Data -num 0 CONN: 1) @ admin-THINK (::1)
WCCILdata    (0), 2017.05.03 12:36:36.287, SYS,  INFO,        4, Connected to (SYS: 1 Data -num 1 CONN: 1) @ admin-THINK (::1)
WCCILdata    (0), 2017.05.03 12:36:36.290, SYS,  INFO,        0, , Manager (SYS: 1 Data -num 1 CONN: 1) initialised
WCCILdatabg  (1), 2017.05.03 12:36:36.456, SYS,  INFO,        6, Initialization by Data Manager finished

[/size]

9 Open System Management -> Database Configuration -> ValArch_201 -> Statistics; click Statistical Update: numbers show that the manager is not working and no data is being archived.

10 However, archive1 statistics show good values.

11. Open para, search for test_archive_201. Open _archive configuration. Simply click "Apply".

12. Repeat step 9: now ValArch_201 statistics show seemingly-good values.

Every time the project is restarted, _archive config for this high-number value archives fails to be used. Re-applying the config in the PARA, solves the problem until next project restart.

Moving the Archive Manager 201 just below the number 5 (before the Event Manager), does not solve the problem.

Starting the Archive Manager 201 with these parameters:

Code: Select all

-num 201 -recv 22 -dbg all 

, produces the lots of log (see attachments)

May be the most interesting part is what appear in the log with these debug settings, after a project restart and just after re-applying changes to _archive config of the test_archive_201 datapoint:

Code: Select all

WCCOAvalarch201:WCCOAvalarch (201), 2017.05.03 13:01:21.921,	INPUTREADY,	(SYS: 1 Data -num 0 CONN: 1) @ WORKSTATION: inputReady
WCCOAvalarch201:WCCOAvalarch (201), 2017.05.03 13:01:21.921,	INPUTREADY,	(SYS: 1 Data -num 0 CONN: 1) @ WORKSTATION: got 316 bytes
WCCOAvalarch201:WCCOAvalarch (201), 2017.05.03 13:01:21.921,	INPUTREADY,	(SYS: 1 Data -num 0 CONN: 1) @ WORKSTATION: All received
WCCOAvalarch201:WCCOAvalarch (201), 2017.05.03 13:01:21.921,	INPUTREADY,	Got message 5 from (SYS: 1 Data -num 0 CONN: 1)
WCCOAvalarch201:WCCOAvalarch (201), 2017.05.03 13:01:21.921,	INPUTREADY,	(SYS: 1 Data -num 0 CONN: 1) @ WORKSTATION: got 0 bytes
WCCOAvalarch201: .921[7352]  write - waitCount:0
WCCOAvalarch201:WCCOAvalarch (201), 2017.05.03 13:01:21.921,	INPUTREADY,	(SYS: 1 Data -num 0 CONN: 1) @ WORKSTATION !receiver.receive_state_all(), only a part of a MSG received. Loop #:2
WCCOAvalarch201: .922[7352]  Time of value points to past! Will be changed to actual one...
WCCOAvalarch201: .922[7352]  The time difference for original time seems to be too big: newTime = 0 (s), lastTime = 1493809106 (s)
WCCOAvalarch201: .922[7352]  Time of new value: 0.1 
WCCOAvalarch201: .922[7352]  Time of previously stored value: 1493809106.29 , it was archiving stop
WCCOAvalarch201:WCCOAvalarch (201), 2017.05.03 13:01:23.019-- Msg overview elapsed secs since last report: 30.005
WCCOAvalarch201:   snd total:      0000000004   rcv total:      0000000006
WCCOAvalarch201:   snd since last: 0000000000   rcv since last: 0000000001
WCCOAvalarch201:-- Msg details
WCCOAvalarch201:man S/R msg                 -
WCCOAvalarch201:-----------------------------------
WCCOAvalarch201:[TOTAL]
WCCOAvalarch201:  Rcv: DP_MESSAGE_VC                   1
WCCOAvalarch201:[(SYS: 0 Data -num 0 CONN: 1)]
WCCOAvalarch201:  Rcv: DP_MESSAGE_VC                   1

https://www.winccoa.com/fileadmin/image ... d_logs.zip

[a.decelis]

Hi,

yes, the managers was already running whilist adding _archive configuration to the DP.