Setup:
Windows 2016 on VM
Disks: San storage
Multiple Oracle 12.1.0.2 patch of Jan 2018 in 2 different oracle homes
1 listener oracle 12.2.0.1 patch of Jan 2018
Why on VM ?
Well, it's the easiest way of setting up a new server to confirm a problem.
Ordering a new physical server and setting it up takes a few weeks.
I've got this serious problem ... well, more or less serious
The listener stops working.
According to some people, and I value their opinion, a listener never stops. period.
Well, I actually have multiple listeners on 2 servers doing that stop thingy. They don't seem to want to follow the ideology of not stopping.
This stopping started on server PHYS TEST304 and now I have the same phenomena on VM TEST310
For the PHYSTEST304 i had opened a case with oracle as multiple database could no longer be connected to, not even with sqlplus / as sysdba, not even with oradebug
after 5 months, i moved to microsoft support, and after several memory dumps, support indicated there were 18 writes waiting to happen in a disk driver. there happened to be 18 databases on that server, coincidence ?
The case with oracle and microsoft never reached a definitve conclusion as we couldn't reproduce the problem at will.
Now on the vm, everything is different, local disks vs san, win2012 vs win2016, physical vs vm
but the oracle base version is the same, all non-cdb 12.1 databases
i've patched all the databases to each version 12.1 from jan 2016 to jan 2018. problem still there.
I was inclined to agree with microsoft that it's the disk driver ...
but now on VM, really ??? same problem ??? different disk driver ... !
not so inclined anymore
Last year, i always found this problem occurring during the oracle maintenance window.
Last night, and the night before, the maintenance window started at 22:00.
And ... the listener had no more entries after 22:00
listener.log
11-JUN-2018 22:01:56 * service_update * W07EU12 * 0
2018-06-11T22:02:00.094998+02:00
11-JUN-2018 22:01:57 * (CONNECT_DATA=(SERVICE_NAME=W10RU12)(CID=(PROGRAM=c:\locusr3+\bin\hostcom_W10R_UPG\hostcom.exe)(HOST=SRVAPSE03SBY)(USER=APSE53))) * (ADDRESS=(PROTOCOL=tcp)(HOST=10.232.8.170)(PORT=60068)) * establish * W10RU12 * 0
2018-06-12T09:05:01.475198+02:00
System parameter file is d:\oracle\ora12201\homeLSNR\network\admin\listener.ora
Log messages written to d:\oracle\ora12201\diag\tnslsnr\vuflorat310\listener\alert\log.xml
Trace information written to d:\oracle\ora12201\diag\tnslsnr\vuflorat310\listener\trace\ora_51988_27144.trc
Trace level is currently 0
Started with pid=51988
Listening on: (DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=vuflorat310.emea.int.grp)(PORT=1521)))
TIMESTAMP * CONNECT DATA [* PROTOCOL INFO] * EVENT [* SID] * RETURN CODE
12-JUN-2018 22:03:35 * (CONNECT_DATA=(SERVICE_NAME=W07EU12)(CID=(PROGRAM=c:\locusr3+\bin\hostcom_W07E_UPG\hostcom.exe)(HOST=SRVAPSE03SBY)(USER=APSE53))) * (ADDRESS=(PROTOCOL=tcp)(HOST=10.232.8.170)(PORT=11433)) * establish * W07EU12 * 0
12-JUN-2018 22:03:35 * service_update * W10RU12 * 0
12-JUN-2018 22:03:35 * service_update * M10RU12 * 0
2018-06-13T08:21:53.701035+02:00
System parameter file is d:\oracle\ora12201\homeLSNR\network\admin\listener.ora
Log messages written to d:\oracle\ora12201\diag\tnslsnr\vuflorat310\listener\alert\log.xml
Trace information written to d:\oracle\ora12201\diag\tnslsnr\vuflorat310\listener\trace\ora_52752_13064.trc
Trace level is currently 0
Started with pid=52752
Listening on: (DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=vuflorat310.emea.int.grp)(PORT=1521)))
TIMESTAMP * CONNECT DATA [* PROTOCOL INFO] * EVENT [* SID] * RETURN CODE
13-JUN-2018 08:21:53 * (CONNECT_DATA=(SERVER=DEDICATED)(SERVICE_NAME=LDB7U12)(CID=(PROGRAM=C:\Program?Files\Unilin\LCP\QAS\Server?Instances\Bazeilles\PrintShop.Server.exe)(HOST=VUFLAPFQ401A)(USER=printshopadmin))) * (ADDRESS=(PROTOCOL=tcp)(HOST=10.232.10.74)(PORT=52874)) * establish * LDB7U12 * 12514
TNS-12514: TNS:listener does not currently know of service requested in connect descriptor
and the maintenance window from the alertlog:
Tue Jun 12 21:41:33 2018
Archived Log entry 673 added for thread 1 sequence 673 ID 0x95a512b dest 1:
Tue Jun 12 22:00:00 2018
Setting Resource Manager plan SCHEDULER[0x30F9]:DEFAULT_MAINTENANCE_PLAN via scheduler window
Setting Resource Manager plan DEFAULT_MAINTENANCE_PLAN via parameter
Tue Jun 12 22:11:34 2018
Thread 1 cannot allocate new log, sequence 675
Private strand flush not complete
Current log# 4 seq# 674 mem# 0: G:\ORACLE\ORADATA\LDBRT12\CNTRL\REDO_004.LOG
Current log# 4 seq# 674 mem# 1: H:\ORACLE\ORADATA\LDBRT12\CNTRL\REDO_054.LOG
Tue Jun 12 22:11:37 2018
Now, my actual question
Taking into account THAT A LISTENER NEVER STOPS ...
Does anybody know why a listener would stop ?
Don't say it doesn't, the proof is above.
I'm looking for reasons, so i can pinpoint the problem
Geert