stderr is getting very large with mutiple entries "NiSelISelectInt"

[Former Member]

Hi all,

during regular system checks we expirienced the file stderrX located in /usr/sap/SID/DVEBMGS00/work/ is getting very large and containes a lot of of identical entries like:

-

---

NiSelISelectInt: 0 handles selected (0 buffered)

Thu Aug 7 14:43:48 2008

NiSelISelectInt: 0 handles selected (0 buffered)

NiSelISelectInt: 0 handles selected (0 buffered)

NiSelISelectInt: 0 handles selected (0 buffered)

NiSelISelectInt: 0 handles selected (0 buffered)

NiSelISelectInt: 0 handles selected (0 buffered)

NiSelISelectInt: 0 handles selected (0 buffered)

NiSelISelectInt: 0 handles selected (0 buffered)

NiSelISelectInt: 0 handles selected (0 buffered)

NiSelISelectInt: 0 handles selected (0 buffered)

Thu Aug 7 14:43:49 2008

NiSelISelectInt: 0 handles selected (0 buffered)

Thu Aug 7 14:43:49 2008

NiSelISelectInt: 0 handles selected (0 buffered)

NiIMyHostName: hostname = 'zosapbw'

NiHsLGetNodeAddr: found hostname 'hostname' in cache

NiIGetNodeAddr: hostname 'hostname' = addr 111.22.3.44

NiIGetServNo: servicename '40000' = port 9C.40/40000

NiICreateHandle: hdl 1 state NI_INITIAL

NiIInitSocket: set default settings for new hdl 1 / sock 8 (UD; ST)

NiIBlockMode: set blockmode for hdl 1 FALSE

NiIInitSocket: set default settings for new hdl 1 / sock 8 (I4; ST)

NiIBlockMode: set blockmode for hdl 1 FALSE

NiICheckPendConnection: connection of hdl 1 to 111.22.3.44:40000 established

NiIConnect: hdl 1 took local address 111.22.3.44:44421

NiIConnect: state of hdl 1 NI_CONNECTED

NiIWrite: hdl 1 sent data (wrt=289,pac=1,MESG_IO)

NiICloseHandle: shutdown and close hdl 1 / sock 8

NiSelISelectInt: 1 handles selected (0 buffered)

NiICreateHandle: hdl 1 state NI_INITIAL

NiIInitSocket: set default settings for hdl 1 / sock 8 (I4; ST)

NiIBlockMode: set blockmode for hdl 1 FALSE

NiIAccept: state of hdl 1 NI_ACCEPTED

NiIAccept: hdl 0 accepted hdl 1 from 111.22.3.44:44422

NiIAccept: hdl 1 took local address 111.22.3.44:40002

NiIRead: hdl 1 received data (rcd=192,pac=1,MESG_IO)

-

---

The SAP systems DB-space is stored on a dedicated SAN (EMC).

Does anyone know the cause of the message and a way how to reduce the frequency of the log writes?

thanks in adavance

T

[Former Member · Accepted solution]

This could be due following reason.

1.You have fiber switch connected to SAN.Get that checked.

2.Heart bit if you have any cluster environment.

3.Do you have IGS ,It write to stderr refer note 1177158

let me know your environemnt.

Regards

Amit

[Former Member]

solution was very simple:

increase igs/tracelevel (RZ10)

stop igs

empty stderr (cat /dev/null > stderr9)

restart igs

verify igs is running and operational

new tracelevel will be active after next application restart

Thanks for support

Thoms