SPARC T3 servers: Oracle Solaris ssh Hangs and System becomes Unresponsive

(Doc ID 1370540.1)

Last updated on AUGUST 17, 2016

Applies to:

Solaris SPARC Operating System - Version 10 10/08 U6 and later
Oracle Solaris on SPARC (64-bit)

Symptoms

sshd "hangs" and will not accept incoming ssh connections.

Changes

The behaviours seen that will point to this problem/solution:

-platform is a Sparc T3 server
-sshd 'hangs' and its processes are unkillable
-following messages are seen in /var/adm/messages: "sshd "monitor fatal: Timeout before authentication for xxx.xxx.xxx.xxx"
-while the sshd hang behaviour is seen, the following command:

echo '::threadlist -v' | mdb -k > threadlist.out

will contain output like this:

00000300ab9bc780 301590c5b60 30160f758a8 1 59 3016277bdb0
PC: cv_wait+0x38 CMD: /usr/lib/ssh/sshd
stack pointer for thread 300ab9bc780: 2a10cff0761
[ 000002a10cff0761 cv_wait+0x38() ]
n2cp_ulcwq_getbuf+0xa8()
n2cp_getreq+0x6c()
alloc_aes_request+0xc()
n2cp_blockinit+0x14c()
common_submit_request+0x630()
process_req_hwp+0xcc()
kcf_submit_request+0xbc()
crypto_cipher_init_prov+0x3a0()
crypto_encrypt_init_prov+0x28()
cipher_init+0x238()
fop_ioctl+0x20()
ioctl+0x184()
syscall_trap32+0xcc()

-ssh/sshd debug output will stop in the following places:
(as seen from the client):

debug1: SSH2_MSG_KEX_DH_GEX_INIT sent
debug1: expecting SSH2_MSG_KEX_DH_GEX_REPLY

(as seen from server):

debug1: set_newkeys: setting new keys for 'out' mode
debug3: aes-128-ctr NID found

 

Cause

Sign In with your My Oracle Support account

Don't have a My Oracle Support account? Click to get started

My Oracle Support provides customers with access to over a
Million Knowledge Articles and hundreds of Community platforms