How to Delete Data in Mango (3.7.7)
-
Greetings, all:
Perhaps I'm alone in this one, but I (sometimes) have data that writes trash into the database. I try to keep it out, but, well, then a sensor breaks lose and is swinging around in the wind writing a bunch of garbagy numbers into something.
The way I used to do this, as described in a previous thread, was to use the Legacy Point Details screen to export (all) the data from the sad point as an excel file. Into excel it goes, with Standard Excel Stuff Applied to find the points to delete. I would then take the modify (add/delete) column, copy'n'paste as values (so the math and stuff doesn't go back into the import file), and save that as an excel, then use the Legacy UI Data Point Import screen to load that back in.
This worked reliably until I upgraded to 3.7.7.. Now, when I do that process, I see the following message:
Error importing row 12, Invalid text in modify column, add or delete only allowedThe modify columns in question, in fact, contain only the word delete or nothing at all. This has worked in the past.
I know with the V4.0 rollout the Legacy UI is going away entirely; Is there still a way to do this sort of "all the data out, filter it, and load it back in" style of feature? Has this function been removed somehow in 3.7.7? Is there a better way to do this?
I realize that I can filter more with metapoints, and set extreme import sets, and lots of other things when the data is first polled into the system. That's all fine and good. But, well, things happen, and sensors go crazy, and sometimes you're left with a big stack of gibberish that needs some hand-adjusting to straighten out. I don't like doing this, since it's a pain in the butt, but, well, sensors do fail in ways that can't easily be filtered on the input.
This particular example relates to a temperature sensor issue: The sensors were damaged somehow, and the values (although in a valid input range) are clearly changing in ways that are not physically possible. We've identified the problem, and fixed it, but I have a reasonable amount of records in the system that need deleting or adjusting to make this data point usable. Any advice here?
I've got a message into customer support: I'll post here when I get a response (unless they respond here first).
Thanks!
-
You can purge in the legacy view for a given timespan, unfortunately it's from days.
Alternatively, if you have swagger enabled... Use the rest API as an admin and delete point values between dates with the XID of the point.
I do agree that 3.7.7 is borked in a variety of ways. So much so that it has cost me a lot of time in firefighting and creating obtuse solutions that are more like absurd temporary fixes.
Hope you're able to get the answers you need.Fox
-
This.. Is a Spicey MeatBall Solution..
So:
(1) The 3.7.7 release has a broken Excel importer, since it's on its way out, anyways.
(2) The only way to modify existing points is to use the REST API as an admin and delete point values using XIDs?
(3) I really don't want to delete.. I want to export, filter, and re-import. I suppose I could do the filtering on a Meta, and "generate history", but there's no way to move that back into the Active polled point, which is used elsewhere in our system for doing other math-related things.I really think there needs to be some way to do this sort of bulk "export / filter / re-import" way native to Mango..
This, and the benefits of Copy and Paste at the data source level, are two things that have been vanished from the Legacy interface that cause me no small amount of consternation.. I mean, the legacy UI could copy'n'paste data sources, which was a really easy way to roll things out. Now I'm exporting JSON, finding and replacing XIDs, and re-importing, which takes at least a billion times as longer than "copy, paste, rename".
Now I have to go whip up another external tool to Mango to do this export / filter / import thing? I mean.. I guess I can write a custom Excel importer using the "data import" thing, and pipe that back into the system... But the "remove" type of excel interface made this so it was a feature we were trying to sell as a benefit to fix this kind of data catastrophe, which happens more often than not.