FP3 supported version of EC2 fleet provider

Hi,

I have also uploaded the requestor configuration for your analysis.

Thanks,

Shruti

Hi,

I uploaded the logs and the questions on the IBM ticket. But if you look at the HF and symA logs, there is a requirement for only 1 slot because there is only 1 task active for 1 application (msgridexec3). But I need multiple SCALE-OUT requests where 6/7 hosts get provisioned and get removed. Why is it so?

Thanks,

Shruti

@Guest Hi Shruti, we didn't make any logic changes when we ported to the V2 API, so if there is an issue, it will remain. It would be better if you clarify what the issue is. "It looks like scale up is not working as expected". I looked at the logs provided and didn't see a problem

Hi Dennis, 

"The current EC2 HF support is based on AWS API V1(which is in EOS), for this RFE, do you need the AWS API V2 support?" 

If the default support for FP3 is AWS API V2, then it would make sense to have awsfleet support on AWS API V2 as well.

Thanks,

Shruti

Hi Dennis, 

"The current EC2 HF support is based on AWS API V1(which is in EOS), for this RFE, do you need the AWS API V2 support?" 

If the default support for FP3 is AWS API V2, then it would make sense to have awsfleet support on AWS API V2 as well.

Thanks,

Shruti

@Guest Hi Shruti, I don't see any provisioning problems. In the symA i can see 2 host being added and released. Where are you seeing issues?

I'm seeing some config issues for symA, I'd need to requestor side conf to figure out why, unless you didn't configure this.
Unable to find return_all_if_zero_tasks. Value of False from Default value is used instead

It looks like scale up is not working as expected. I have attached the logs, could you help to figure out the issue and help suggest a fix?

Hi,

Yes I am testing FP3. I removed the lines and tested it, however, I see a weird behavior. I have uploaded the logs on the IBM Ticket for troubleshooting.

Thanks, Shruti

@Guest Hi Shruti, are testing in FP3? If so we removed these lines from all the scripts, since these didn't work after upgrading to java 17. Approx line 31-41.

<! if [ "$_java" ]; then
<! JAVA_VER=$("$_java" -version 2>&1 | sed -n ';s/.* version "(.).(.)..*"/\1\2/p;')
<! #version=$("$_java" -version 2>&1 | awk -F '"' '/version/ {print $2}')
<! #echo version "$JAVA_VER"
<! if [ "$JAVA_VER" -lt "18" ]; then
<! echo "Java version error. HostFactory: AWSFleet provider plugin requires Java version 1.8 or up"
<! exit 1
<! fi
<! fi

Shruti,

The current EC2 HF support is based on AWS API V1(which is in EOS), for this RFE, do you need the AWS API V2 support?

Thanks, Dennis

I got the below error whilst testing the awsfleet provider - 

2026-09-28 10:36:01.377 UTC TRACE [200960:139753305843584] hf_dumpInputOutputFiles: Output json is: /var/tmp/symphony/kernel/conf/../../hostfactory/conf/providers/awsfleetinst/scripts//getAvailableTemplates.sh: line 36: [: 170 2025-04-15: integer expression expected { "templates" : [ {    "templateId" : "EC2-Fleet-Template",    "maxNumber" : 312,    "attributes" : {      "nram" : [ "Numeric", "4096" ],      "ncpus" : [ "Numeric", "16" ],      "ncores" : [ "Numeric", "1" ],      "type" : [ "String", "X86_64" ]    },    "instantMode" : false,    "launchTemplateVersion" : "$Default",    "launchTemplateABSEnabled" : false,    "launchTemplateReservedCapacityEnabled" : false,    "launchTemplateSubnetId" : null,    "launchTemplateImageId" : null,    "onDemandCapacity" : 0 } ], "message" : "Get available templates success." 
 

I fixed the script locally and ran into another issue - 

​[2026-09-29 16:40:45.512 UTC ERROR com.ibm.spectrum.util.AwsFleetUtil.requestEC2Fleet(AwsFleetUtil.java:1045)] Failed to request an EC2 fleet: com.amazonaws.services.ec2.model.AmazonEC2Exception: Parameter: OnDemandOptions.AllocationStrategy: prioritized and OnDemandOptions.MaxTotalPrice can not be specified together. (Service: AmazonEC2; Status Code: 400; Error Code: InvalidParameterCombination; Request ID: 97c76f98-9260-498b-8a9c-a6e96e030374; Proxy: null)        at com.amazonaws.http.AmazonHttpClient$RequestExecutor.handleErrorResponse(AmazonHttpClient.java:1879)        at com.amazonaws.http.AmazonHttpClient$RequestExecutor.handleServiceErrorResponse(AmazonHttpClient.java:1418)        at com.amazonaws.http.AmazonHttpClient$RequestExecutor.executeOneRequest(AmazonHttpClient.java:1387)        at com.amazonaws.http.AmazonHttpClient$RequestExecutor.executeHelper(AmazonHttpClient.java:1157)        at com.amazonaws.http.AmazonHttpClient$RequestExecutor.doExecute(AmazonHttpClient.java:814)        at com.amazonaws.http.AmazonHttpClient$RequestExecutor.executeWithTimer(AmazonHttpClient.java:781)        at com.amazonaws.http.AmazonHttpClient$RequestExecutor.execute(AmazonHttpClient.java:755)        at com.amazonaws.http.AmazonHttpClient$RequestExecutor.access$500(AmazonHttpClient.java:715)        at com.amazonaws.http.AmazonHttpClient$RequestExecutionBuilderImpl.execute(AmazonHttpClient.java:697)        at com.amazonaws.http.AmazonHttpClient.execute(AmazonHttpClient.java:561)        at com.amazonaws.http.AmazonHttpClient.execute(AmazonHttpClient.java:541)        at com.amazonaws.services.ec2.AmazonEC2Client.doInvoke(AmazonEC2Client.java:39468)        at com.amazonaws.services.ec2.AmazonEC2Client.invoke(AmazonEC2Client.java:39435)        at com.amazonaws.services.ec2.AmazonEC2Client.invoke(AmazonEC2Client.java:39424)        at com.amazonaws.services.ec2.AmazonEC2Client.executeCreateFleet(AmazonEC2Client.java:4753)        at com.amazonaws.services.ec2.AmazonEC2Client.createFleet(AmazonEC2Client.java:4722)        at com.ibm.spectrum.util.AwsFleetUtil.requestEC2Fleet(AwsFleetUtil.java:1035)        at com.ibm.spectrum.awsfleet.AwsFleetImpl.requestMachines(AwsFleetImpl.java:451)        at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method)        at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(Unknown Source)        at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(Unknown Source)        at java.base/java.lang.reflect.Method.invoke(Unknown Source)        at com.ibm.spectrum.AwsFleetMain.call(AwsFleetMain.java:169)        at com.ibm.spectrum.AwsFleetMain.main(AwsFleetMain.java:603) [2026-09-29 16:40:45.772 UTC ERROR com.ibm.spectrum.util.AwsFleetUtil.requestEC2Fleet(AwsFleetUtil.java:1045)] Failed to request an EC2 fleet: com.amazonaws.services.ec2.model.AmazonEC2Exception: Parameter: OnDemandOptions.AllocationStrategy: prioritized and OnDemandOptions.MaxTotalPrice can not be specified together. (Service: AmazonEC2; Status Code: 400; Error Code: InvalidParameterCombination; Request ID: c902c23d-9852-4eea-af6f-d4c0dbecd7f1; Proxy: null) 
 

I removed the parameter - "maxTotalPriceOnDemandPerRequestPerHour": 20 , from the template as it is incorrectly configured, and it moved past the error to submit the fleet request. However, there are other errors encountered in the fleet scripts.

2026-09-30 10:27:13.551 UTC TRACE [2925759:140148739618688] hf_dumpInputOutputFiles: Output json is: /var/tmp/symphony/kernel/conf/../../hostfactory/conf/providers/awsfleetinst/scripts//requestMachines.sh: line 36: [: 170 2025-04-15: integer expression expected { "message" : "Request VM success from AWS.", "requestId" : "fleet-7137b32e-a92c-e63c-ae12-272abe5acd3c" } 2026-09-30 10:27:13.551 UTC TRACE [2925759:140148739618688] ExecuteAction::parseOutput: Output json FAILS here 2026-09-30 10:27:13.551 UTC WARN [2925759:140148739618688] SubDemand::triggerAlloc: No output from requestMachines 2026-09-30 10:27:13.557 UTC TRACE [2925759:140148739618688] HFDBUtils::updateOneFieldRecordInDB(): Execute sql (update hf_request set status='Done with Errors' where id='sd-4c1ec8e0-a84c-462a-823a-7d19d6ae7fbe') successful. 2026-09-30 10:27:13.557 UTC TRACE [2925759:140148739618688] DemandGeneric::incrementNumSubdemandsDone: Demand set to state Provisioned

Another issue is that when the demand is killed, the retrunrequest.sh script doesn’t return any hosts to retrun resulting into capacity being maintained in the cluster.

2026-09-30 10:40:05.684 UTC TRACE [2925759:140148632831744] ExecuteAction::spawnNonBlocking: Execute command:/var/tmp/symphony/kernel/conf/../../hostfactory/conf/requestors/symA/scripts//getReturnRequests.sh -f /tmp/hf_in_savX4f > /tmp/hf_out_pjGb2h 2>&1 2026-09-30 10:40:06.086 UTC TRACE [2925759:140148632831744] hf_dumpInputOutputFiles: Input json is: { "demand_requests_pending": [], "demand_requests_inprogress": [], "allocated_hosts": [], "relinquished_hosts": [], "deallocated_hosts": [] } 2026-09-30 10:40:06.086 UTC TRACE [2925759:140148632831744] hf_dumpInputOutputFiles: Output json is: {"return_hosts": []}