800xA Information Management trends freezes

Hello,

We have an issue with the IM hierarchical trend, from time to time the trend freezes, but the data collected is OK. What could be causing it? I'm attaching some images. System800xA version 6.0.3.3.

Because of that, the following happens:
- When the trend i shown in seamless in 2 hours interval, the data is shown OK.
- When the trend is shown in seamless in 8 hours interval, the data gets distorted.
- This only happens with the trends that have IM logs configured.

Best regards,

Comments (0) New comment

alvaro.ruizdesomocurcio Rank: 850

Asked 9 months ago

Answer this question Follow

Answers

Please provide screen shots from the Log Configuration aspect showing:

1. Data collection tab for the OPC Log

2. Data Collection tab for the IM Log

There can be some issues if not configured correctly.

Comments (0) New comment

OursLars Rank: 188

Answered 9 months ago

Looks like a bug - please use formal support channels to initiate a support case so that we can engage R&D and start working on a solution.

Comments (0) New comment

Stefan Stromqvist AKS expert

Rank: 1

Answered 9 months ago

Dear Alvaro,

What you are seeing may indeed be a bug, and if so it should be fixed. However, over the years I have seen several examples of incorrect Log Template configuration showing up the way you described. If you don't mind, please provide screen shots of the configuration of the OPC an IM logs in the template, showing how data is collected, processed and stored. If this looks correct, please raise a support case. If the configuration is wrong, I can let you know how to fix it.

Comments (0) New comment

OursLars Rank: 188

Answered 9 months ago

Hi, Here is the information related to the original template. I've already created a new one to test, because one parameters seems to be incorrect (sample blocking rate).

Here is the new template created (IM definition and deadband are still the same).

Greetings,

Comments (7) New comment

Stefan Stromqvist AKS expert

Rank: 1 on 10/12/2020 6:30:50 AM | Like (1) | Report

1)
Do *NOT* use "Store As Is" if the hierarchical log has a different Storage Interval than the primary log's Min Time. Use e.g. "Average". Or else, the hierarchical log will likely wrap prematurely since "it stores all samples read", even though it expect "less". Log period will end too fast.

Sample Blocking Rate is too high.
1000 minutes is almost 1 day. I suggest using the "hsdbmaint -stagger" tool to adjust SBR down to 1 hour.

Please apply these two changes and revert with the outcome.

Stefan Stromqvist AKS expert

Rank: 1 on 10/12/2020 6:39:05 AM | Like (1) | Report

Stagger is detailed in 3BUF001092-610 A, chapter 13.14.11
In principle, run "hsdbmaint -stagger" in an elevated CMD.EXE prompt, answer "y" to continue, "y" to use defaults and "y" to stagger now. Then stop/start IM History services using PASGUI.exe

C:\>hsdbmaint -stagger
hsRS.c @ 367 | File mapping created mapHandle = 24800000000
hsRSGetSegment @ 364 | Attach to segment 0
-----------------------------------------------------------------------------
Collection Info Per Controller: ( Rates are in units per minute )
-----------------------------------------------------------------------------
Ctrl: Dev Sub # Logs Sample Rate Request Rate
----------- --- --- ------ ----------- ------------
0(OPCHDA) 0 0 79 154.0 1.7
-----------------------------------------------------------------------------
Stagger Summary Information: ( Sample/Storage/Blocking units are seconds )
-----------------------------------------------------------------------------
Total(Type) Time Sample Storage Blocking Range AvgRate
------------------ ------------------ ------ ------- -------------- ---------
24(OPC HDA) 01 Jan 90 05:00:00 15 60 1800/1800 0.80
18(OPC HDA) 01 Jan 90 05:00:00 60 300 3600/3600 0.30
1(OPC HDA) 01 Jan 90 05:00:00 15 600 3600/3600 0.02
36(OPC HDA) 01 Jan 90 05:00:00 60 600 3600/3600 0.60
------------------ ------------------ ------ ------- -------------- ---------
Average Requests Per Minute from TTD/PHL: 0.00
Average Requests Per Minute from OPC HDA: 1.72
Average Requests Per Minute to hsStorage: 1.72
-----------------------------------------------------------------------------

Do you wish to continue? [yn] y
Default Summary:
24(OPC HDA) logs with blocking 30:00( 120 points ), stagger = 30
18(OPC HDA) logs with blocking 60:00( 60 points ), stagger = 12
1(OPC HDA) logs with blocking 30:00( 120 points ), stagger = 3
36(OPC HDA) logs with blocking 60:00( 60 points ), stagger = 6
Average requests per minute from TTD/PHL: 0.00
Average requests per minute from OPC HDA: 1.73
Average requests per minute to hsStorage: 1.73

Do you wish to use the defaults? [yn] y

Do you wish to stagger NOW? [yn] y
hsStagger.c @ 1403 | 1. START 'Stagger Init '...
24(OPC HDA) elms, stagger 30, subGroup 1, blocking 1800 ( 30 min )
hsStagger.c @ 1459 | 1. STOP 'Stagger Init '- 24, time = 00:00:01 or 1 sec
hsStagger.c @ 1403 | 1. START 'Stagger Init '...
18(OPC HDA) elms, stagger 12, subGroup 1, blocking 3600 ( 60 min )
hsStagger.c @ 1459 | 1. STOP 'Stagger Init '- 18, time = 00:00:00 or 0 sec
hsStagger.c @ 1403 | 1. START 'Stagger Init '...
1(OPC HDA) elms, stagger 3, subGroup 1, blocking 1800 ( 30 min )
hsStagger.c @ 1459 | 1. STOP 'Stagger Init '- 1, time = 00:00:00 or 0 sec
hsStagger.c @ 1403 | 1. START 'Stagger Init '...
36(OPC HDA) elms, stagger 6, subGroup 1, blocking 3600 ( 60 min )
hsStagger.c @ 1459 | 1. STOP 'Stagger Init '- 36, time = 00:00:01 or 1 sec

In the above example, stagger chose SBR to 30 or 60 minutes (depending on the log's Sample Interval).

Then just use PASGUI.exe to first "Stop All" and then "Start All"

by OursLars Rank: 188 on 10/12/2020 7:15:15 AM | Like (0) | Report

Well, first of all: What he (Stefan) said. This is indeed along the lines of what I suspected: If you don't have an calculation applied (Average, max...) to the data in the IM Log, the sample rate of the IM log must match the storage rate of the OPC log or you will see problems like what you have seen. The data in the IM log is not even the correct data for the time. What I tend to do is not to use the IM deadband but only the Exception Deviation in the OPC with a small value (but greater than zero). The IM Log would have the same sample and storage interval as the OPC log, but with with more data points (or longer time).

The OPC log may have have a minimum time of x seconds and a maximum time of x times a factor, as per your new log. You could have a storage of a couple of weeks instead of just five days. Most requests would then just be serviced from the OPC Log. The IM Log should then look as per your example and you will get a storage depth of much more than 104 weeks unless for very noisy signals.

Your original set-up is not unusual - I have had to fix this kind of problem several times over the years.

by alvaro.ruizdesomocurcio Rank: 850 on 10/12/2020 7:28:41 AM | Like (0) | Report

Hi, I ran the steps detailed above, but the problem is that in the System I can't change any values of the data collection without having to erase all the configured logs and lose all the data already saved.

Stefan Stromqvist AKS expert

Rank: 1 on 10/12/2020 7:54:52 AM | Like (1) | Report

Bottom line, your configuration is invalid and you must do something about it.

"hsdbmaint -stagger" will overwrite SBR on a database level but you cannot change Calculation Algoritm this way, only in template which is only possible before creating a log referencing it.

Also, see this thread: https://forum-controlsystems.abb.com/20232262/Log-configuration?tc=ca2e58c7-89d3-403b-a3ad-0db153ee46fa

by OursLars Rank: 188 on 10/12/2020 4:41:04 PM | Like (0) | Report

Dear Alvaro, you are NOT losing any real data if you reconfigure (i.e. re-create) the IM Log. What is in it now is rubbish or 'fake data': The values stored that appear to be for a certain time are not from that time. The longer you wait correcting the IM Log configuration, the more 'fake data' you will have. I say the IM Log, because you can keep the OPC Log configuration as long as the sample rate of the new IM Log matches the storage rate for the OPC Log. In other words, you should be able to remove the IM Log from the template and then create a new, correct IM Log in the template. This will be pushed out to all the instances of the template.

by OursLars Rank: 188 on 10/12/2020 6:13:22 PM | Like (0) | Report

Just a tip if you are going to replace the IM Log in the template: Make sure that after deleting the IM Log that it gets deleted on all instances before you add the new one.

Add new comment

alvaro.ruizdesomocurcio Rank: 850

Answered 9 months ago

Question details

800xA Information Management trends freezes

Answers

Other questions needing answers

Related topics

AKS experts