Query Monitoring: delete from system:prepared where RequestId = <some existing ID> is inconsistent on windows, same test passes on centos fails on windows

Test which passes on centos but fails on windows:

 python testrunner.py -i rqg.ini -c conf/tuq/py-tuq-monitoring.conf  -p doc-per-day=1

What the test does?

creates prepared statements and checks in select * from system:prepareds.

Gets a particular requestID and then deletes from resultset of system:prepareds and then checks if it actually got deleted and doesnt appear in resultset.

Please edit conf/tuq/py-tuq-monitoring.conf  to have:

tuqquery.newtuq.QueryTests:

    test_simple_check,cbq_version=sherlock,dataset=sabre,monitoring=True

    test_joins_monitoring,cbq_version=sherlock,dataset=sabre,monitoring=True

    test_simple_nulls,cbq_version=sherlock,dataset=sabre,monitoring=True

    test_limit_offset,cbq_version=sherlock,dataset=sabre,reload_data=False,monitoring=True

    test_simple_alias,dataset=sabre,reload_data=False,monitoring=True

    test_alias_from_clause,dataset=sabre,reload_data=False,monitoring=True

    test_alias_order_desc,dataset=sabre,monitoring=True

tuqquery.tuq.QueryTests:

    test_named_prepared_between,named_prepare=prepare_between,doc-per-day=1,monitoring=True

test_named_prepared_between fails on windows but passes on centos.

I debugged through the tests and found that after deleting a particular requestId from system:prepareds, it still exists and hence resultcount of prepared statements is not 0 in the beginning of next test on windows while on centos the the particular requestId gets deleted quickly and the  resultcount of prepared statements is 0 in the beginning of next test.

Example of ini file:

global]

port:8091

username:Administrator

password:Membase123

index_port:9102

[servers]

1:SERVER_1

[SERVER_1]

ip:10.17.1.34

index_port:9102

port:8091

services=n1ql,kv,index

[membase]

rest_username:Administrator

rest_password:password

Please use your own windows box.