Slow Oracle Net Client Connection Caused by DNS Issue (Doc ID 2242222.1)

Last updated on MARCH 31, 2022

Applies to:

Symptoms

Tnsping response times are very high.  SQL*Plus connections via the listener (TCP) may become slow as hostname resolution is taking a lot of time.

Connections to the database via TCP listener are slow, and SQL*Net trace for shadow process shows snlinGetAddrInfo step spends time (please see snippet below)

20-JAN-2017 01:05:46:985] snlinGetAddrInfo: entry
20-JAN-2017 01:05:46:985] snlinGetAddrInfo: getaddrinfo() failed with error -2
20-JAN-2017 01:05:46:985] snlinGetAddrInfo: exit
20-JAN-2017 01:05:46:985] nttbnd2addr: looking up IP addr for host: VIP.com
20-JAN-2017 01:05:46:985] snlinGetAddrInfo: entry
20-JAN-2017 01:05:47:352] snlinGetAddrInfo: exit
20-JAN-2017 01:05:47:352] snlinFreeAddrInfo: entry

.

20-JAN-2017 01:05:47:356] nsdo: nsctx: state=2, flg=0x4005, mvd=0
20-JAN-2017 01:05:47:356] nsdo: gtn=10, gtc=10, ptn=10, ptc=32730
20-JAN-2017 01:05:47:356] nscon: entry
20-JAN-2017 01:05:47:356] nscon: recving a packet
20-JAN-2017 01:05:47:356] nsprecv: entry
20-JAN-2017 01:05:47:356] nsprecv: reading from transport...
20-JAN-2017 01:05:47:356] nttrd: entry

Note the gap in time here is greater than 3 minutes.

20-JAN-2017 01:08:58:798] nttrd: socket 18 had bytes read=8
20-JAN-2017 01:08:58:798] nttrd: exit==================================================================>connection exit  after 3 minutes @ 01:08:58. Still no response from DNS 
20-JAN-2017 01:08:58:798] nsprecv: 8 bytes from transport
20-JAN-2017 01:08:58:798] nsprecv: tlen=8, plen=8, type=11
20-JAN-2017 01:08:58:798] nsprecv: packet dump

Cause

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.