In transaction code SM50, there are lots of work processes are stuck due to ENQ reason and the report status shown as SAPLSENA.
In the trace file of the impacted work process, you will see below error messages,
***LOG Q0I=> NiIRead: P=192.168.1.10:3201; L=192.168.1.20:49963: recv (10054: WSAECONNRESET: Connection reset by peer) [nixxi.cpp 5
*** ERROR => NiIRead: SiRecv failed for hdl 12/sock 1376
(SI_ECONN_BROKEN/10054; I4; ST; P=192.168.1.10:3201; L=192.168.1.20:49963) [nixxi.cpp 5098]
{root-id=005056825628A9F803E31AIJKC0EE}_{conn-id=7B6198E92836EDH6881C7F}_1
In dev_enqlisten (work directory of ASCS), below error message captured,
Thr 11860] Sat Apr 11 07:20:35 2020
[Thr 11860] *** WARNING => NiICreateHandle: no more free handles (6014) [nixxi.cpp 464]
[Thr 11860] *** WARNING => NiICreateHandle: no more free handles (6014) [nixxi.cpp 464]
[Thr 11860] *** WARNING => NiICreateHandle: no more free handles (6014) [nixxi.cpp 464]
Solution
Referring to SAP Note 2196513 – ENSA threads with: “NiICreateHandle: no more free handles (XXXX)” or “maximum of sockets reached” errors, you need to increase the value of parameter enque/server/max_clients.
The value is changeable from the ASCS profile parameter and not in the DEFAULT or INSTANCE profile.
There are some cases where there are non SAP work processes are accessing to the ENQ port like i.e. anti-virus software or monitoring tools software. This will bring huge traffic accessing the same port of ENQ server.
