From: Software Groups (sfgroups@gmail.com)
Date: Mon Aug 07 2006 - 23:29:26 EDT
Looks like problem with one host, when batch jobs running, it losses
the network connecting, form that server I was not able to ping
default gateway.
Here is netstat s command output, Is this output looks normal?
RAWIP
rawipInDatagrams = 5247 rawipInErrors = 0
rawipInCksumErrs = 0 rawipOutDatagrams = 4887
rawipOutErrors = 0
UDP
udpInDatagrams =669682 udpInErrors = 0
udpOutDatagrams =700570 udpOutErrors = 0
TCP tcpRtoAlgorithm = 4 tcpRtoMin = 400
tcpRtoMax = 60000 tcpMaxConn = -1
tcpActiveOpens =387486 tcpPassiveOpens =2236130
tcpAttemptFails = 4052 tcpEstabResets = 28710
tcpCurrEstab = 658 tcpOutSegs =3686501853
tcpOutDataSegs =3565412936 tcpOutDataBytes =3371633015
tcpRetransSegs =13935950 tcpRetransBytes =2783222380
tcpOutAck =117283172 tcpOutAckDelayed =16590179
tcpOutUrg = 0 tcpOutWinUpdate = 2949
tcpOutWinProbe = 2939 tcpOutControl =5247874
tcpOutRsts = 7436 tcpOutFastRetrans =10390254
tcpInSegs =2247106769
tcpInAckSegs =1882098886 tcpInAckBytes =3905074533
tcpInDupAck =91856682 tcpInAckUnsent = 0
tcpInInorderSegs =976875420 tcpInInorderBytes =2970183904
tcpInUnorderSegs = 19513 tcpInUnorderBytes =23708915
tcpInDupSegs =117640 tcpInDupBytes =8101130
tcpInPartDupSegs = 584 tcpInPartDupBytes =444224
tcpInPastWinSegs = 19 tcpInPastWinBytes =1832077384
tcpInWinProbe = 0 tcpInWinUpdate = 2932
tcpInClosed = 120 tcpRttNoUpdate =2792622
tcpRttUpdate =1877052564 tcpTimRetrans =20154403
tcpTimRetransDrop = 2235 tcpTimKeepalive =16408674
tcpTimKeepaliveProbe=6915641 tcpTimKeepaliveDrop = 741
tcpListenDrop = 1032 tcpListenDropQ0 = 0
tcpHalfOpenDrop = 0 tcpOutSackRetrans = 52308
IPv4 ipForwarding = 2 ipDefaultTTL = 255
ipInReceives =2053070322 ipInHdrErrors = 0
ipInAddrErrors = 0 ipInCksumErrs = 0
ipForwDatagrams = 0 ipForwProhibits = 0
ipInUnknownProtos =8831687 ipInDiscards = 0
ipInDelivers =2244655136 ipOutRequests =3507582785
ipOutDiscards = 8828 ipOutNoRoutes = 113
ipReasmTimeout = 60 ipReasmReqds = 26
ipReasmOKs = 26 ipReasmFails = 0
ipReasmDuplicates = 0 ipReasmPartDups = 0
ipFragOKs = 26 ipFragFails = 0
ipFragCreates = 71 ipRoutingDiscards = 0
tcpInErrs = 0 udpNoPorts =1359609
udpInCksumErrs = 0 udpInOverflows = 0
rawipInOverflows = 0 ipsecInSucceeded = 0
ipsecInFailed = 0 ipInIPv6 = 0
ipOutIPv6 = 0 ipOutSwitchIPv6 = 810
Thanks
On 8/7/06, Software Groups <sfgroups@gmail.com> wrote:
> Hi Managers,
>
> We have about 50 Solaris servers with NIS, autofs setup. We have this
> setup working fine about 5 years.
>
> Now the problem is on few servers, we are getting "NIS server not
> responding" error message. This happens only from 10pm to 3am, the
> problem lost about 10 minutes to 2 hours, after that all services
> works fine. During daytime we are not having any problem.
>
> All server running Solaris 8, NIS master server patch level
> Generic_117350-27. NIS slave server patch level Generic_117350-33
>
> One of the NIS client server, we ran shell script to ping other
> server; during NIS problem time we got 100% ping package loss. We
> opened ticket with network team; they checked the Cisco switch.
> 1. They didn't find any error message on the switch
> 2. Servers connected using VLAN
> 3. No firewall between servers.
>
> Some servers we have two network interface, one for normal network
> other one for backup network, ifconfig command show same mac address
> for both interface (hme0, hme1). Network team said, it might be issue
> with different mac address.
>
> 1. In syslog I didn't see any IP address conflict or network
> connection UP/DOWN error message.
> 2. eeprom shows local-mac-address?=false
> 3. Server load is normal during that time.
> 4. NIS master server uptime is 183 days, client server uptime 30 to 50
days
>
> I am looking for some troubleshooting steps, to resolve this issue.
>
> 1. Is this related to server patch level?
> 2. Why this problem occurs only in the night, at random time.
> 3. These servers running for years, last couple of months we install
> onlu OS patche on these severs. Other than patch nothing changes on
> this configuration.
> 4. If it is network switch issue, if they way to capture some debug
> message on host side?
>
> Thanks
>
>
> --
> Software Groups (SFG)
> http://www.sfgroups.com
>
-- Software Groups (SFG) http://www.sfgroups.com _______________________________________________ sunmanagers mailing list sunmanagers@sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers
This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 23:40:33 EDT