I am testing vectorsearch, while doing so I am trying to paginate my resultset for a "next page" function to give me the first, second, third 15 entries within a table. 

For this I have two embedding classes. One with a HNSW Index (vectornomicembedtextlatest) , and one without (vectornomicembedtexttest).

Calling SELECT ID,PRIMKEY FROM SQLUser.vectornomicembedtexttest LIMIT 5 OFFSET 1 works fine with the first entry having the rowID of 486448. (I deleted old entries in the beginning and reused the table)

SELECT ID,PRIMKEY FROM SQLUser.

Windows 10/Windows Server 2016:

I am currently monitoring our license use with a new rest-service I am implementing when I noticed my licenses on my instance being consumed and never released by Visual Studio Code.

For this I restarted my Instance, watched the licenses for a while, which remained at 1 during idle (I am guessing my MMGT Portal session uses 1) 
But when I connect to my Instance using Visual Studio Code (with my Instance setup in the extension already), suddenly 2 licenses are used.

This may be a very specific problem but might help with some general information that I wasn't able to find anywhere.

I have started to train 2 IntegratedML models using AutoML one month ago.

My first attempt was very optimistic with a class that has  26 Properties and 100.000 datasets. I wasn't expecting good results, just wanted to test it. 
(CREATE MODEL ECLASSPREDICT3 PREDICTING (eClass) FROM CRMSHOPARTIKELML)
My second attempt was more specific using the same class structure but reducing the dataset to 10.000 entries and only using certain properties.

When trying to create a ML Model I ran into the problem, that AutoML apparantly isn't available on my IRIS instance.

I was able to use the following command successfully: CREATE MODEL ECLASSPREDICT PREDICTING (eClass) FROM SQLUSER.CRMSHOPARTIKEL

Then when trying to use TRAIN MODEL ECLASSPREDICT, I get the error from the image:

Roughly translates to : "-186: Modelprovider not available on this instance. #2822 %ML provider 'AutoML' is not available on this instance"

In my ML Configurations there are 3 standard configurations called %AutoML, %H20 and %PMML which I haven't changed.

After installing IRIS 2023.1 on a live copy of our production machine our REST Service now consumes a CSP Session with every request. The request is handled as expected, but uses one of the 5 CSP Session per license. So after 25 requests, the license is used up. The Grace time always shows 0 and the session stay for very long. (Maybe the 900 Seconds timeout). 
On Caché 2018, we had the same settings for the Webapp and there, only a single Session was set for all requests. The Caché request didn't use any cookies. 
Here we also used  $system.License.

I'm having trouble adjusting the protocol used for any CSP Requests. Currently all request to the system are sent over HTTP 1.1

When loading 40+ images on a page, the browser starts stalling request. According to google HTTP 1.1 only allows 6 parallel TCP Connections before suspending the remaining. 

I have looked through the Web Gateway Settings, %Net.HttpRequest.HTTPVersion and SSL Configuration but haven't found anything.

Is there a hidden setting that can be changed? Does Caché even support higher HTTP protocols. We do have an IRIS development instance I could test on.

In the Windows Ressource Manager I can observe multiple parallel processes coming from cache.exe with read operations to journaling files.

All except one of these processes have the same reads(Byte/s). The processes point to different journal files and constantly read between 200 and 3000 Bytes/s.

The corresponding process via PID in the management portal of Caché shows the process %SYS.Monitor.Control.1. In 3 days of uptime on the server it has run 181.632.583 commands and modified 32.140.642 globals. 

These processes are not writing to the journal at any point.