Timeseries doesn't load in ETL

[SbiCA]

I'm struggling to get a timeseries loaded.

I followed the documentation https://ravendb.net/docs/article-page/6.0/csharp/server/ongoing-tasks/etl/raven#time-series

loadToEventDocuments(this);


function loadCountersOfEventDocumentsBehavior(documentId, counterName) {
   return true;
}

function loadTimeSeriesOfEventDocumentsBehavior(docId, timeSeriesName) {
   // only normal ts
   if(timeSeriesName =="AppOpen"){
       return true;
   }
   return false
}

my collection is called EventDocuments and I have an ETL setup just for this one collection with the following script for what ever reason the time series AppOpen doesn't appear on the target.

This I tried to make the change

1. reset ETL

2. recreate ETL

3. delete the existing document on the target:

   • change source doc
   • reset ETL to load doc

Also to give the complete context
Source server is a raven cloud instance running 5.4.112
Target server is self hosted instance running 5.4.112

I'm pretty much clueless

[aviv86]

Hi

This is a bug.. apparently we are passing the lower time series name as the timeSeriesName argument.

I created a ticket to fix that, you can track it here:
https://issues.hibernatingrhinos.com/issue/RavenDB-21866/ETL-TimeSeries-load-behaviour-function-passes-lower-timeseries-name-as-argument

As a workaround, you can use:

function loadTimeSeriesOfEventDocumentsBehavior(docId, timeSeriesName) {
   // only normal ts
   if (timeSeriesName =="appopen"){
       return true;
   }
   return false
}

Regards

[SbiCA]

Thanks for looking into it @aviv86

I can confirm after switching the script it worked for me but for what ever reason the suggested lower casing didn't to the trick I switched to the following.

function loadTimeSeriesOfEventDocumentsBehavior(documentId, timeSeriesName) {
   // only normal ts
   if(!timeSeriesName.startsWith("inc")){
       return true;
   }
   return false
}

[SbiCA]

I'm wondering if this shouldn't become the default anyway since incremental time series are not supported yet

see #15193