Federation

6 Topics

Digital twins and large data quantities - what are your performance tips to deal with millions of (federated) datasets?
Hi community, Aras currently posts a lot of digital twin related content and we jumped on the hype train. We right now combine EBOM/MBOM with ERP, production and test data. User can see data of individual products or can do all kinds of reporting over the complete data. We are still at the beginning of this concept. Right now people are mainly interest in reports and calculations. It´s nothing we haven´t done before. Certain data was already federated in Innovator in the past. But users had to connect certain data from the various sources manually or ask an admin to create reports that are more sophisticated. The digital twin concept allows more freedom to approach. But the bottleneck is currently the test data. We do automated tests of electronic components. A few hundreds tests per individual product are common. So the amount of data is very huge (millions+). Certain queries and reports that are currently possible might probably fail in the future due to the increasing amount of data. It´s not an issue right now, but I want to be prepared for the future. What are the best practices to optimize performance regarding large amounts of federated data? I checked the official Platform Specification that is available online. It contains many hints regarding hardware recommendations for a certain amount of users, Vaults, etc. It´s a useful guide, but it doesn´t cover federated data and digital twin use cases yet. What do you recommend? What can we do on database / infrastructure level? 1. Optimize database structures (better query design, indexing, caching, etc.)? 2. Throw more hardware at the problem? This one is not necessarily an Innovator issue. PLM and SQL are designed to deal with a huge amount of data. There are big corporations which thousands of users and endless data vaulted over the whole world. But I would be happy about anyone that can give some additional insights, ideas or recommendations! Many thanks!! Angela
AngelaIp
7 months ago Place Development
0Views
1like
2Comments
Best way to highlight (fully) federated data in grids and views?
Hi community, does anyone has some ideas how to highlight federated data? I have some ItemTypes that show data from other data sources. The data is not mixed with Innovator data, it´s complete external content. I added a few customizations so also paging and search bar works like for regular items. This work so well that users right no often cannot tell that the data doesn´t come from Innovator. Isn´t that a good thing? Yes! But no! Cause people expect they can work with the data like with any other. But federated data has some limitations: - Federated views are sometimes slower regarding loading time (often due to query complexity) - While most UI elements work, some still doesn´t (like the calculate max pages button) -- EDIT: Max page works now, but not stuff like "sort across pages" - Users cannot query for all data at once cause of hard max page restrictions - The data cannot be edited inside Innovator - Sometimes it´s not live data, but a few hours old My idea is to use add something that indicates that data comes from an external data source. This way we wouldn´t mess up with user expectation which result in better game play (this is how a game designer would describe it). But I right now not sure what's the best way to highlight Federated ItemTypes. There are three scenarios: - Federated Search Grid - Federated Form - Federated Relationship Grid Do you have some ideas how or where to add some hint that the data is not from Innovator? I use the "Federated" flag in the ItemType. But it´s not big help. Thanks for any input! Angela
Solved
AngelaIp
2 years ago Place Development
0Views
0likes
1Comment
Federating Files in ARAS
Hello, I am currently looking into federating some data coming from another system in ARAS. The objects themselves are already being displayed correctly but now I'm looking into being able to display the main file for each of there objects in ARAS. I wanted to ask if there is a way to "federate" the files while still being able to use the native ARAS file item fields, etc... The amount of files is quite high so if possible I would like to avoid creating File Items for all of them... Thank you, Bernardo
BernardoP
6 years ago Place Development
2.5KViews
1like
1Comment
How to synchronize the data of MPP_process plan with all its releationships with PostgreSQL Database
Hi, I'm a research student and we are doing a university project on PLM, we have an idea of storing and interacting the data of PLM systems with other systems like MES and ERP so we thought to create a central database through which every system will interact with other systems. Now we have to synchronize the data of MPP_process plan with all its releationships with PostgreSQL Database where the central DB is located, we wanna do this as a trail test to sync the data from Aras server into our Central Database and vice-versa. We are thinking not only to export the data for single time, but to Auto update of the data atmost real time. One of my friend from this Aras community suggested me to do by federation services but I don't know how to apply and how it works.! Or you have any other ideas how we can do? someone suggest me how to do. Thanks inadvance.
venkatsai
7 years ago Place Development
5.2KViews
0likes
0Comments
How to color full federated items in the grid?
Hi, maybe somebody has an idea how to solve this one. I have a full federated ItemType that contains a couple of yes/no columns. The data is displayed as standard grid in Aras. I want to highlight the yes/no values in the grid. Coloring regular ItemTypes is possible with an onAfterGet Method that overwrites the CSS property. There are samples available in this forum that show how to do it. But this approach doesn´t work for full federated Items, as the physical table inside Aras is empty and we just lock at datasets from another database. I already discovered, that I can fill the Aras 'CSS' property in the SQL-view that I use for quering the data. But unfortunately, my current SQL view is very complex. I would have to do a lot of SQL queries for building the correct stylesheet string for the CSS property column. Does anybody know a simpler approach to solve this use case? Thanks for any feedback! Angela
Solved
AngelaIp
7 years ago Place Development
6.2KViews
0likes
3Comments
Federation - Item count
Hello, I've created a federated ItemType to pull in data from an ODBC data source and it seems to be working great apart from one thing: When running a search the status bar at the bottom reports '0 items found', and consequently you can't page through the results. 25 results do actually appear on screen though! Looking at the results in AML Studio I can't see any difference between what my onGet returns and the results of any other query on a normal native ItemType, so I'm not sure how I'm supposed to communicate the number of results (and page number) to the UI. To save uploading all my code, my results are assembled by: Create inn.newItem("my federated type") Set attributes and properties Get AML of new Item and append to string Repeat this for all results Strip the isNew and isTemp attributes from the AML Wrap in <Result> tags Return via inn.newItem().LoadAML(mystring); This is probably the most roundabout way of assembling a result set but I couldn't find any better way that worked at least as far as I've got. Any pointers to documentation on how to do this better, or ideas on how to handle the item count gratefully received. Many thanks in advance, Ben ps. We're on V11 SP5
benjamin_brooking
9 years ago Place Development
2.7KViews
0likes
2Comments