My Flow seems to be stucked in a loop with Sharepoint updateitem

[cfoenki]

Hi everybody,

I'm struggling with an issue on my flow.

Basically my flow has for input a json table that I need to pass to a Sharepoint list either as an update when the element exists or a new item when the element is new.

As I cannot determine which items changed since the last call on my json table, I need to update all items each time.

It is working well when I have a loop of around 270 items (completed in 7 minutes)

When I look at my other loop for a little bit more than 1000 items it is taking forever (still ongoing)

I took a look on the associated sharepoint list and it seems that nothing is happening for the last 20 minutes (8 hours ago is my previous run)

I would be extremely greatful if you have a solution.

Thank you in advance!

[v-chengfen-msft]

Hi @cfoenki ,

Maybe you can try to enable the concurrency function of [Apply to each]

This can improve the execution efficiency of [Apply to each] and reduce the execution time

Best Regards

Cheng Feng

[cfoenki]

@v-chengfen-msft Thank you for the response. Unfortunately turning on Concurrency control made things worse 😞

[grantjenkins]

Are you able to show what you have in your Condition within the loop?

Also, does your List and JSON data have a unique property that you can match on (a way to identify which JSON object relates to which List item)?

If you could show a sample of your JSON data and the List item, that would help.

[cfoenki]

Sure, here is what is inside my loop

I have 3 parrallel loops for 3 different sharepoint list could it cause the issue?

I tested to save my flow as a new one and keep only one branch and it is working fine

[grantjenkins]

I'd suggest explaining exactly what your flow is trying to achieve (including the three SharePoint Lists), and provide a sample of the JSON data and the List data. This would help to perhaps redesign the flow so it's more efficient.

Also a screenshot(s) of the entire flow as you have it now.

[cfoenki]

Thank you for taking the time to answer. I redesigned the flow and now it is working but maybe you could help me having a more efficient flow as mine is taking about half an hour.

I have an Excel table on a Sharepoint document library. There are 3 tables of interest in that Excel: Deviation, CAPA and Other CAPA Plan.

My excel file has no Table objects (and I cannot change that)

From the Excel file, I want to create 3 sharepoint lists that are mirror from the Excel Table data. I have then 3 lists:
- Deviation Table Dashboard --> Should contain data from "Deviation" tab of my Excel

- CAPA Plan Table Dashboard --> Should contain data from "Other CAPA Plan" tab of my Excel

- CAPA Table Dashboard --> Should contain data from "CAPA" tab of my Excel

I want that my Sharepoint lists are updated regularly with Excel file updates

My Sharepoint lists are linked to a PowerApps application.

I built an office script to extract data to power automate.

Here is the script:

/**
 * This script returns the values from the used ranges on each worksheet.
 */
function main(workbook: ExcelScript.Workbook): WorksheetData[] {
    // Create an object to return the data from each worksheet.
    let worksheetInformation: WorksheetData[] = [];

    // Get the data from every worksheet, one at a time.
  workbook.getWorksheets().forEach((sheet) => {
    // Check if the sheet name is different from "Audit Trail"
    if (sheet.getName() !== "Audit Trail") {
      let values = sheet.getUsedRange()?.getValues();
      worksheetInformation.push({
        name: sheet.getName(),
        data: values as string[][]
      });
    }
  });

  return worksheetInformation;
}

// An interface to pass the worksheet name and cell values through a flow.
interface WorksheetData {
    name: string;
    data: string[][];
}

I use sheet.getUsedRange to detect the range of data in each sheet.

In Power Automate I use the Run script action to retrieve output

The array result has this kind of format (I trimmed data a lot so it is not too long):

{
	"body": [
		{
			"name": "Deviation",
			"data": [
				[
					"Reference",
					"Process",
					"Service",
					"Creator",
				],
				[
					"#####",
					"ZZZZ",
					"YYYYY",
					"XXXX"
				],
				[
					"#####2",
					"ZZZZ2",
					"YYYYY2",
					"XXXX3"
				]
			]
		},
		{
			"name": "CAPA",
			"data": [
				[
					"ReferenceCAPA",
					"ProcessCAPA",
					"ServiceCAPA",
					"CreatorCAPA",
				],
				[
					"#####",
					"ZZZZ",
					"YYYYY",
					"XXXX"
				],
				[
					"#####2",
					"ZZZZ2",
					"YYYYY2",
					"XXXX3"
				]
			]
		},
		{
			"name": "Other CAPA Plan",
			"data": [
				[
					"ReferenceCAPAPlan",
					"ProcessCAPAPlan",
					"ServiceCAPAPlan",
					"CreatorCAPAPlan",
				],
				[
					"#####",
					"ZZZZ",
					"YYYYY",
					"XXXX"
				],
				[
					"#####2",
					"ZZZZ2",
					"YYYYY2",
					"XXXX3"
				]
			]
		}
	]
}

As I cannot use it like this I use the filter action to first select deviation table results:

And then a Select action to convert the result to a JSON array

The output is like this (with my previous example):

{
	"body": [
		{
			"Reference": "Reference",
			"Process": "Process",
			"Service": "Service",
			"Creator": "Creator"
		},
		{
			"Reference": "#####",
			"Process": "ZZZZ",
			"Service": "YYYYY",
			"Creator": "XXXX"
		},
		{
			"Reference": "#####2",
			"Process": "ZZZZ2",
			"Service": "YYYYY2",
			"Creator": "XXXX2"
		},
		{
			"Reference": "#####3",
			"Process": "ZZZZ3",
			"Service": "YYYYY3",
			"Creator": "XXXX3"
		}
	]
}

To remove the header row I use a Compose action with skip function:

Then I use a Parse JSON action to easily retrieve my fields:

Then I retrive all the current items in the sharepoint list (using Pagination as I may have a lot of items)

And finally I have a loop through all items of my JSON table

In the loop I first make a filter on the output of get items to check if reference is foundin the existing list

If length is equal to 0, it means that the item is new, I used then a Create Item action to fill a new line, if it is >0 then I used an Update Item action to update the existing line

The formula inside the field is checking if the value of the field is N/A or with no text, if so field is filled with null value otherwise it is using the Current Item property corresponding

If(or(equals(items('Loop_through_all_deviation_lines')['Reference'],''),equals(items('Loop_through_all_deviation_lines')['Reference'],'N/A')),null,items('Loop_through_all_deviation_lines')['Reference'])

The formula in the Id field of Update Item is related to the previous filter function:

first(body('Filter_on_the_current_deviation_item'))?['ID']

And that is the end, except that this is repeated 3 times as I have 3 tables.

Here is the screenshot of the whole flow for 1 list:

If you have an idea on how to optimize this please let me know 🙂

Thank you in advance.

[grantjenkins]

What you have is pretty well constructed given what you're working with (Excel file with three tables that aren't defined as Tables, etc.). I'm assuming the main bottleneck is the loops due to the Condition and Add/Update actions.

A few questions.

1. How many items would be in the three tables (approximately).
2. How many items would be in the existing SharePoint Lists?
3. How many would be new items vs. existing items that needed to be updated (approximately)?
4. Are the Reference values always unique (what you are using in your Condition)?
5. Do you have access to Premium connectors?

[cfoenki]

@grantjenkins Your are right the bottleneck is the update/createitem loop.

After some tests on my existing flow the flow takes 30-40 min to run when trigger is manuel but when it is a trigger Schedule with a récurrence of 1 per day, the flow is taking forever (when I write this it is still ongoing after 7 hours...)

Here are the answers to the questions:

1. How many items would be in the three tables (approximately).

In Déviation I have roughly 300 items

In Other CAPA Plans 50 items

In CAPA 1000 items

1. How many items would be in the existing SharePoint Lists?

Same as in Excel (mirror)

1. How many would be new items vs. existing items that needed to be updated (approximately)?

New items for Deviation approximately 5 per week Updates 50 items per week

New items for Other CAPA Plan 2 per week Update 1 item per week

New items for CAPA 20 per week Update 200 per week

1. Are the Reference values always unique (what you are using in your Condition)?

Yes all references are unique

1. Do you have access to Premium connectors?

Yes

[RADical]

@cfoenki You might be hitting an infinite loop in the action. You can try changing that connection to use a different user or create a condition which updates the list when the flow is triggered.

Preventing Infinite Loop in MS Power Automate (MS Flow) - YouTube

[grantjenkins · Accepted solution]

Hopefully this will speed things up significantly. It checks if you have new items and adds them, and checks if any of the items were updated, and only updates those ones. So, it will only loop for new or updated items.

For this example, I've got the following Excel file with three worksheets. The ones highlighted in yellow are an item that's been updated, and a new item.

And the following SharePoint Lists. Note that I've renamed the Title field to Reference in this example.

See full flow below (for one parallel branch). I'll go into each of the actions. Note that this includes all the actions for Deviations branch. The other two branches would be essentially the same.

Run script is the same as what you already have.

Get items Deviation returns all the items from the list. I've set Top Count to 5000, Pagination On, and a Threshold of 10000. This will retrieve items in batches of 5000 (instead of the default 100 which is slower), up to 10000 items.

Filter array Deviation is the same as what you have where it filters out the Deviation object from the Office Script results.

Select Excel Deviation is the same as what you have to extract out the properties. The only change is that I'm stripping out the first row (headers) within the From expression.

//From
skip(first(body('Filter_array_Deviation'))?['data'], 1)

//Reference, Process, Service, Creator
item()[0]
item()[1]
item()[2]
item()[3]

Parse JSON Deviation is the same as what you have. The Schema used is:

{
    "type": "array",
    "items": {
        "type": "object",
        "properties": {
            "Reference": {
                "type": "string"
            },
            "Process": {
                "type": "string"
            },
            "Service": {
                "type": "string"
            },
            "Creator": {
                "type": "string"
            }
        },
        "required": [
            "Reference",
            "Process",
            "Service",
            "Creator"
        ]
    }
}

Items to be added

Firstly, I work out what items need to be added (don't exist in the list). To do this, I get all the References from the list, then check what items in the Excel results are not contained within that - those are the new items we need to add.

Select Deviation References retrieves a simple array of all References from the list.

Filter array Deviation To Be Added filters out the data so only items that need to be added are retrieved. From is using the output from Parse JSON Deviation. Output comes from Select Deviation References. Reference gets the Reference from the current item.

//From
body('Parse_JSON_Deviation')

//Output
body('Select_Deviation_References')

//Reference
item()['Reference']

Items to be updated

Next, I work out what items have been updated (difference in either Process, Service or Creator). I'm using some XPath expressions here.

XML Deviation converts our data from Select Excel Deviation to XML so we can use XPath.

xml(json(concat('{"root": { value:', body('Select_Excel_Deviation'), '}}')))

Select Combined Deviation combines the existing data in our SharePoint List Items with the corresponding data from Excel based on the Reference (Title field in my example). It uses the following XPath expressions.

//ProcessXL
xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference="', item()?['Title'], '"]/Process/text())'))

//ServiceXL
xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference="', item()?['Title'], '"]/Service/text())'))

//CreatorXL
xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference="', item()?['Title'], '"]/Creator/text())'))

Filter array Deviation To Be Updated uses the output from Select Combined Deviation and the following filter expression, which will check if any of the values have been updated in Excel.

@OR(
    not(equals(item()?['Process'], item()?['ProcessXL'])),
    not(equals(item()?['Service'], item()?['ServiceXL'])),
    not(equals(item()?['Creator'], item()?['CreatorXL']))
)

Parse JSON Deviation To Be Updated uses the output from Filter array Deviation To Be Updated, and the following Schema.

{
    "type": "array",
    "items": {
        "type": "object",
        "properties": {
            "ID": {
                "type": "integer"
            },
            "Reference": {
                "type": "string"
            },
            "Process": {
                "type": "string"
            },
            "Service": {
                "type": "string"
            },
            "Creator": {
                "type": "string"
            },
            "ProcessXL": {
                "type": "string"
            },
            "ServiceXL": {
                "type": "string"
            },
            "CreatorXL": {
                "type": "string"
            }
        },
        "required": [
            "ID",
            "Reference",
            "Process",
            "Service",
            "Creator",
            "ProcessXL",
            "ServiceXL",
            "CreatorXL"
        ]
    }
}

Apply to each Add Deviation iterates over each of the items from Filter array Deviation To Be Added. And within the loop, it uses Create item Deviation to add each of the new items.

Apply to each Update Deviation iterates over each of the items from Parse JSON Deviation To Be Updated. And within the loop, it uses Update item Deviation to update each of the items that have changed.

In my example, both loops will only iterate once, since I've only got one item that was updated and one item that was added. I would hope the entire flow (with all your data) would complete within a couple of minutes all going well.

The output in my Deviations Table Dashboard List after the flow runs is below.

[cfoenki]

Hi @grantjenkins ,

Thank you for this step by step explanation.

I followed your explanation and scaled it to my real data model (48 columns).

I got an error on the step "Select Conbined Deviation":

The execution of template action 'Select_Combined_Deviation' failed: The evaluation of 'query' action 'where' expression '{
  "ID": "@item()?['ID']",
  "Reference": "@item()?['Title']",
  "Process": "@item()?['Process']",
  "Service": "@item()?['Service']",
  "Creator": "@item()?['DeviationCreator']",
  "Detection date": "@item()?['Detectiondate']",
  "Creation date": "@item()?['DeviationCreationdate']",
  "Responsible": "@item()?['DeviationResponsible']",
  "Type": "@item()?['DeviationType']",
  "Procedure Reference": "@item()?['ProcedureReference']",
  "Description": "@item()?['Description']",
  "Immediate actions": "@item()?['Immediateactions']",
  "Transmitted to Compliance on": "@item()?['TransmittedtoComplianceon']",
  "Initial Regulatory risk": "@item()?['InitialRegulatoryrisk']",
  "Initial Regulatory risk comment": "@item()?['InitialRegulatoryriskcomment']",
  "Initial Product/Patient risk": "@item()?['InitialProduct_x002f_Patientrisk']",
  "Initial Product/Patient risk comment": "@item()?['InitialProduct_x002f_Patientrisk0']",
  "Initial Deviation level": "@item()?['InitialDeviationlevel']",
  "Initial Justification": "@item()?['InitialJustification']",
  "Date PR (for critical)": "@item()?['DatePR_x0028_forcritical_x0029_']",
  "Date EUQPPV (for critical)": "@item()?['DateEUQPPV_x0028_forcritical_x00']",
  "Evaluation Date": "@item()?['EvaluationDate']",
  "Due Date": "@item()?['DueDate']",
  "Investigation Conclusion": "@item()?['InvestigationConclusion']",
  "Root cause identified": "@item()?['Rootcauseidentified']",
  "Root cause analysis": "@item()?['Rootcauseanalysis']",
  "Root cause category": "@item()?['Rootcausecategory']",
  "Recurrent Deviation": "@item()?['RecurrentDeviation']",
  "Recurrence analysis": "@item()?['Recurrenceanalysis']",
  "Perimeter extension": "@item()?['Perimeterextension']",
  "New Perimeter of impact": "@item()?['NewPerimeterofimpact']",
  "Perimeter extension comment": "@item()?['Perimeterextensioncomment']",
  "Investigation conclusion date": "@item()?['Investigationconclusiondate']",
  "Final Regulatory risk": "@item()?['FinalRegulatoryrisk']",
  "Final Regulatory risk comment": "@item()?['FinalRegulatoryriskcomment']",
  "Final Product/Patient risk": "@item()?['FinalProduct_x002f_Patientrisk']",
  "Final Product/Patient risk comment": "@item()?['FinalProduct_x002f_Patientriskco']",
  "Final Deviation level": "@item()?['FinalDeviationlevel']",
  "Final Justification": "@item()?['FinalJustification']",
  "Final PR Notification date": "@item()?['FinalPRNotificationdate']",
  "Final EUQPPV notification date": "@item()?['FinalEUQPPVnotificationdate']",
  "Deviation closure date": "@item()?['Deviationclosuredate']",
  "CAPA needed": "@item()?['CAPAneeded']",
  "CAPA Plan reference": "@item()?['CAPAPlanreference']",
  "CAPA Plan Creation Date": "@item()?['CAPAPlanCreationDate']",
  "CAPA Plan Objective": "@item()?['CAPAPlanObjective']",
  "CAPA Plan Responsible": "@item()?['CAPAPlanResponsible']",
  "CAPA Plan due date": "@item()?['CAPAPlanduedate']",
  "CAPA Plan closure date": "@item()?['CAPAPlanclosuredate']",
  "ProcessXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Process/text())'))",
  "ServiceXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Service/text())'))",
  "CreatorXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Creator/text())'))",
  "Detection dateXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Detection date/text())'))",
  "Creation dateXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Creation date/text())'))",
  "ResponsibleXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Responsible/text())'))",
  "TypeXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Type/text())'))",
  "Procedure ReferenceXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Procedure Reference/text())'))",
  "DescriptionXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Description/text())'))",
  "Immediate actionsXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Immediate actions/text())'))",
  "Transmitted to Compliance onXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Transmitted to Compliance on/text())'))",
  "Initial Regulatory riskXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Initial Regulatory risk/text())'))",
  "Initial Regulatory risk commentXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Initial Regulatory risk comment/text())'))",
  "Initial Product/Patient riskXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Initial Product/Patient risk/text())'))",
  "Initial Product/Patient risk commentXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Initial Product/Patient risk comment/text())'))",
  "Initial Deviation levelXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Initial Deviation level/text())'))",
  "Initial JustificationXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Initial Justification/text())'))",
  "Date PR (for critical)XL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Date PR (for critical)/text())'))",
  "Date EUQPPV (for critical)XL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Date EUQPPV (for critical)/text())'))",
  "Evaluation DateXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Evaluation Date/text())'))",
  "Due DateXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Due Date/text())'))",
  "Investigation ConclusionXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Investigation Conclusion/text())'))",
  "Root cause identifiedXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Root cause identified/text())'))",
  "Root cause analysisXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Root cause analysis/text())'))",
  "Root cause categoryXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Root cause category/text())'))",
  "Recurrent DeviationXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Recurrent Deviation/text())'))",
  "Recurrence analysisXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Recurrence analysis/text())'))",
  "Perimeter extensionXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Perimeter extension/text())'))",
  "New Perimeter of impactXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/New Perimeter of impact/text())'))",
  "Perimeter extension commentXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Perimeter extension comment/text())'))",
  "Investigation conclusion dateXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Investigation conclusion date/text())'))",
  "Final Regulatory riskXL": "@xpath(ou

Are you able to help me on that one? I am wondering it it could not be a matter of space in my Initial JSON table column names

[grantjenkins]

Yes, it looks like it's because you've got spaces in your Field Names that you're using in your XPath expressions. You'll need to use the converted field names for any that have spaces.

//If you had the following Field Name

First Name

//It would be converted to

First_x0020_Name

//Spaces will be replaced with

_x0020_ 

[cfoenki]

Thanks I will test that

[cfoenki]

I ended up by removing all spaces in my columns but still get the error (here is the simplified select action for test):

InvalidTemplate. The execution of template action 'Select' failed: The evaluation of 'query' action 'where' expression '{ "ID": "@item()?['ID']", "Reference": "@item()?['Title']", "Process": "@item()?['Process']", "ProcessXL": "@xpath(outputs('XML_Deviation'), concat('string(//root/value[Reference=\"', item()?['Title'], '\"]/Process/text())'))" }' failed: 'The template language function 'xpath' failed to parse the provided XML.'.

Edit: I saw that despite the checkmark, the step of XML creation is returning an error on the output:

After further investigation the above error is due to a special character

[cfoenki]

Hi @grantjenkins ,

I finally figured out what to do with my special characters embbed in the Excel text. I modified my office script to replace all special characters by a space and now after several attemps it is WORKING!! Using your method.

Now my flow takes only 20 minutes and moreover it is working with the same speed using recurrence trigger.

Thank you so much for the help!!

[takolota]

Could this all be done in a simpler & faster way using SharePoint batch updates?

https://powerusers.microsoft.com/t5/Power-Automate-Cookbook/Batch-Update-SharePoint-List/td-p/136541...