Agent Down

HI
I have oracle 10g database installed on Redhat Enterprise AS 4.about 4 days ago we had a power outage and the the server went down for about 8 hours.after I started the mechine I've found that the agent is not working properly and no way to run applications:(
I tried so many ways on working with netca,netmgr,lsnrctl.gradually i solved the problem by adding the loopback IP address in /etc/hosts and shuting down the emctl and recreating a new listener.
now after 2 days The agent went down for about 50 minutes.and started automatically withour my intervention.
Is there any suggestion about the causes of Agent down?what makes agent to goes down ?is it possible that anyting works with listener may be corrupt because of that power outage.
Many Thanks In Advace

OK the problem is solved...
I do not know WHY there was a problem after the upgrade but in frustration we just went to the OEM console and pulled up from "all Targets" the agent on that box and dropped and then re-added the database instance to the list of "monitored targets" and up it came.
Thanks for your attention to this question...

Similar Messages

EM Event: Warning: Database Instance - Agent Down

Hello,
while upgrading plugins for the OMS 12 Bundle Patch and applying the BP on the agents, I get a lot of mails with the above subject - namely, one mail by every database instance and every listener running on the host.
Content of the mail is as follows:
Host=<hostname>
Target type=Database Instance
Target name=<database name>
Categories=Availability
Message=Agent Down
Severity=Warning
Availability status=Agent Down
Rule Name=Database Availability and Critical States,Target Availability Event Rule
Rule Owner=SYSMAN
Blackout on the host leads to the OMS not being able to do the upgrades - obviously a blacked-out agent is not able to do that. What puzzles me that in the Rule mentioned in the mail, incidents should only be created for the target type being down (i.e. for database instance when database is down, and so on) as one would expect.
I would rather not blackout all targets on a host EXCEPT the agent as this gets old with our number of hosts and dbs, does anyone know which rule is responsible for the mails and how they could be suppressed temporarily in an efficient way?

I think Jay was referring to selecting availability events by severity. So in the rule create/edit "Select Events" step you can do the following:
1. select (event) Type= Target Availability
2. Leave the radio button in the sub-section at "All events of type Target Availability"
3. select the check box next to "Severity" right below the Type field
4. When a subsequent drop down shows up for severity - select "Fatal"
5. Optionally, check the box next to "Target type" and choose "Database Instance" if your target set includes other targets - this step should not be needed.
6. Everything else should remain the same.
Essentially, the "target down" events come with "Fatal" severity and these "agent down" events come with a "Critical" severity. So this is a way to avoid getting the agent down notifications on databases (or any other targets) while getting the genuine target down notifications. Hope this helps.

Agent down not reported in OEM

I am on 11.1.0 OEM grid control. One of the cluster host's agent is down. But the same is not reported on the Pie chart of OEM home page.
When i click on the Host performance tab, it reports the below error.
An error has occurred!
Unable to obtain data for target hostA. The target may be down. Switching to the last 24 hrs view.But for the RAC instance running on this node, OEM does not report any error. It gives all details as if the agent is up. (even after waiting for 10 min)
In the same OEM when the agent on any other target is down, it is reported in the pie chart within 5 min.
I am not clear what is wrong with this and how to fix it.

Hi,
I am getting following error when adding database manually
oracle.sysman.emSDK.emd.comm.CommException: Received unexpected response text : <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML><HEAD><TITLE>Error Message</TITLE> <META http-equiv=Content-Type content="text/html; charset=UTF-8"> <STYLE id=L_default_1>A { FONT-WEIGHT: bold; FONT-SIZE: 10pt; COLOR: #005a80; FONT-FAMILY: tahoma } A:hover { FONT-WEIGHT: bold; FONT-SIZE: 10pt; COLOR: #0d3372; FONT-FAMILY: tahoma } TD { FONT-SIZE: 8pt; FONT-FAMILY: tahoma } TD.titleBorder { BORDER-RIGHT: #955319 1px solid; BORDER-TOP: #955319 1px solid; PADDING-LEFT: 8px; FONT-WEIGHT: bold; FONT-SIZE: 12pt; VERTICAL-ALIGN: middle; BORDER-LEFT: #955319 0px solid; COLOR: #955319; BORDER-BOTTOM: #955319 1px solid; FONT-FAMILY: tahoma; HEIGHT: 35px; BACKGROUND-COLOR: #d2b87a; TEXT-ALIGN: left } TD.titleBorder_x { BORDER-RIGHT: #955319 0px solid; BORDER-TOP: #955319 1px solid; PADDING-LEFT: 8px; FONT-WEIGHT: bold; FONT-SIZE: 12pt; VERTICAL-ALIGN: middle; BORDER-LEFT: #955319 1px solid; COLOR: #978c79; BORDER-BOTTOM: #955319 1px solid; FONT-FAMILY: tahoma; HEIGHT: 35px; BACK
Thanks,

Emctl clearstate with agent down.

We had a few targets that had alert log alerts stuck within Grid and would not clear, despite that the errors had been remedied and after over 24 hours.
I stopped the agent then issued clearstate then restarted and fine - all cleared.
However, in the book Oracle Enterprise Manager 10g Grid Control Handbook, we are told:
"never issue a clearstate command when the Agent is down" If it is not up and running when the command is issued the state of the targets will not be reset".
This does not reflect my experience.
Does anyone have thoughts on this?
Thanks,
DA.
Oracle Enterprise Manager 10g Release 5 Grid Control 10.2.0.5.0

OK the problem is solved...
I do not know WHY there was a problem after the upgrade but in frustration we just went to the OEM console and pulled up from "all Targets" the agent on that box and dropped and then re-added the database instance to the list of "monitored targets" and up it came.
Thanks for your attention to this question...

In 12c OEM where can I set the notification when the Agent is down.

Hello Experts,
In 10g OEM environment , while creating the notification rule , we have a tab "availablity" which has the option to choose for the Agent target availability.
But while creating a rule in the 12c , I didn't find that kind of option. So, how can I enable my self to alerted for the agent nonavailability.
Thanks,
Naveen kumar

EM12c has an entirely different set up for notifications. You would configure this under the Setup menu by selecting Incidents, followed by Incident Rules.
Within Incident Rules you should see a couple default Oracle-provided incident rulesets. I created an additional ruleset (priority 3, after the default rulesets) that contains various rules, each rule matching a type of incident for which I want to receive notification.
Specifically for the "agent down" situation, my rule is set up as:
Target Type is Agent
Status in (Closed; New)
Category is Availability
Then you specify actions to occur when the rule gets triggered -- I've specified the action to email my EM administrator login, which is configured with my email address.

Except JSP Agent, other web components are down.

In my R12 Vision instance, Except JSP Agent, other web components are down. How can i make it up & running? Please help me.
Web Components Status
PL/SQL Agent Down
Servlet Agent Down
JSP Agent Up
Discoverer Unmonitored
Personal Home Page Down
TCF Down
Thanking you

As applmgr user, please do the following:
- cd %INST_TOP%\apps\<SID>_<hostname>\admin\scripts
- Issue "adopmnctl.sh start" to startup the components
- Issue "adopmnctl.sh status" to verify the status of each component, you should get something similar to the following:
Checking status of OPMN managed processes...
Processes in Instance: PROD_db2.db2.domain.com
-------------------+--------------------+---------+---------
ias-component      | process-type       |     pid | status
-------------------+--------------------+---------+---------
OC4J               | oafm               |    3800 | Alive
OC4J               | forms              |    2528 | Alive
OC4J               | oacore             |    3112 | Alive
HTTP_Server        | HTTP_Server        |     172 | Alive
ASG                | ASG                |     N/A | Down
exiting with status 0
ERRORCODE = 0 ERRORCODE_END

Agent Connection failure

I got a Database Instance and agent down.
Agent Connection to Instance
Status Failed
Details ORA-12505: TNS:listener does not currently know of SID given in connect descriptor (DBD ERROR: OCIServerAttach)
But Listener is UP
I check tnsping and ping . all works. I use oracle10g R2 in window 2003
I try to startup or recovery.
I enter database sys or dab id/password and remote PC ID/password.
But I got a error message for my password. Actually, I can use these ID/PW to remote PC and access database.
I am new.
What Do I need to do for this condition?
Thanks
JIm

Thanks for you help. plz see copy message.
1)
Agent Connection to Instance
Status Failed
Details ORA-12505: TNS:listener does not currently know of SID given in connect descriptor (DBD ERROR: OCIServerAttach)
2)database instance
Database Instance
Port 1521
SID sale
Oracle Home D:\oracle\product\10.2.0\sale
3)Listener
Listener
Status Up
Port 1521
Name LISTENER
Oracle Home D:\oracle\product\10.2.0\sale
Location D:\oracle\product\10.2.0\sale\network\admin
I made ping and tnsping.and lrsnctl. i do not find issue.
I am waiting for you help in detail.
Nice day
JIm

Stopped Blackout but host/agent remain in blackout state

I created a blackout on a host and all targets and set it to be indefinite. Worked fine. Finished what I wanted to do and wanted to start monitoring again. went to Setup / Blackout and selected to Stop the blackouts. Now no more blackout targets are showing from Setup/Blackouts page. But, my targets still show blacked out more than an hour later. The agent is running fine on the target host, can upload, etc, but if I look at the host target in GC, the status says "agent down", the icon shows a little blue circle with a black down arrow. I think normally if the agent is "down" I see only a red down arrow. I think this icon is indicating I'm still in blackout. I can't get it to come back. Ideas?

Guess I should have thought of that. I figured since it uploaded ok and the emctl status agent showed it was uploading metrics, etc that stop/start wouldn't do anything. just for good measure, re-ran the secure agent as well and now it works.
thanks.

WLS Server shows as down, but is up in WLS console

I have a WLS Domain discovered in OEM (12c). I used the same agent configuration and setup on all 3 servers in the cluster. Server 1 and 2 show as up, as well as the cluster and overall domain. Server 3 shows as down with numerous incidents flagged with Agent down and Weblogic Server Is Down. In WLS Console, all three servers are Running and working properly.
What sort of emctl commands can be utilized to verify that the agent on Server 3 is communicating properly with the WLS? Or steps within OEM, perhaps to verfiy the communication. I have checked config files and setup procedures, and Server 3 is identical to Server 1 and 2, aside from the 1/2 vs 3 and corresponding hostnames. Looking for any sort of input...

That is why I am confused. The agent is actually up. It is secured and sending metrics for the box perfectly. I can perform ./emctl upload agent successfully. When I navigate to Host Targets, the box shows as up. But when I navigate to Middleware Targets, on the WLS domain, Server 3, which is on that box, shows as down. The agent unreachable is cleared on that page, but it shows the WebLogic Server as down.
When I access the WLS console, the server is up, running and working fine. That is why I am lost when it comes to my next step. If agent is working fine and the setup is identical to server 1 and 2. Why is it down, when the rest are up? I noticed in the Server3.log file, I am getting spammed with a Certificate error. "CERTIFICATE_UNKNOWN alert was recieved from *Server 3*. The peer has an unspecified issue with the certificate." I am getting the cert warning from server 3 on server 3. So is there a problem with some certificate and that is why the agent on server 3 can not communicate with the WLS?

Grid is still showing Tagert after uninstalling the agent

Hi DBAs,
I installed the agent (10.2.0.4.0) on Linux_64 (RHEL) box and configured it with Grid Control. I was able to see the target in the Grid Conrol. I need to uninstall the agent from this host in order to configure it with Other Grid Control. Now once I uninstalled the agent from the target. I am still seeing the target host in grid with agent down status. I want to remove the target from the grid, plz let me know how can I do that.
Regards
-Samar-

Hi Samar,
Ideally its recommended to first remove the target from Grid Control console as mentioned in NOTE.271691.1. As you have already uninstalled the agent, issue the following command connected as sysman user from repository database to remove the agent and targets:
conn sysman
exec mgmt_admin.cleanup_agent('agentname:port');
commit;
The above command removes the agent and all targets montiorred by the agent.It might take time to remove the target depending on the number of targets monitored.
Hope this helps.
--Lai

Updating Linux agents to latest code - GW 8.02 HP2

I've been asked to update our POA on the Linux box to 8.02 HP2 to troubleshoot some issues we're having with the Data Synchronizer.
I was told that updating the agents on Linux was much easier than on Netware, in that you could simply update the files, and stop/restart the processes, thus minimizing any downtime (whereas with Netware, you had to take the agents down before you could even begin the process).
Is this correct? For example, I can run the update, choose INSTALL PRODUCTS | GROUPWISE AGENTS | INSTALL AGENTS to copy the updated code. Is it then a matter of simply running "rcgrpwise restart" to load the new code? Or must I run the CONFIGURE AGENTS first?
This doesn't seem to be very well-documented, and this is the first time I've had an opportunity to update the Linux version of Groupwise, but of course, it's hard to get downtime, too, and my Data Synchronizer system is broken until I can get my POA updated.
Any suggestions?
Thanks!

You are correct. Currently at 8.02 -- simply needing the SOAP fixes in HP2. I am only running a POA on this box.
In fact, I had already run INSTALL AGENTS to see what options there were, only to be surprised when it began copying files and completed with no further prompts!?!
Since I now know I don't need to run CONFIGURE AGENTS, I'm debating whether to initiate the "rcgrpwise restart" to load the new code (and possibly disrupt service if problems arise), or wait until the middle of the night, which will delay resolving my Data Syncronizer problem yet another day...

Photos Agent draining battery quickly!

I have a late 2014 13" retina Macbook which, since updating my software to be able to carry 'photos' instead of iPhoto, the battery has been extremely hot and losing energy very quickly because of 'Photos Agent'. I checked out activity monitor, and also tried to quit photos agent, but the threads manage to reappear after a few seconds, and it is using around 90% of my battery and I need a way to reliably shut PHOTOS AGENT down and make sure it doesn't reappear. Apple have made a mistake with this as it is extremely inconvenient and unnecessary. Thank you for any help!

Open the iCloud preference pane and uncheck the box marked Photos, if it's checked.

Problem in deleting backup job in Grid Control

Hi,
The topology and versions are as follows:
1 windows 2003 server with Oracle Grid Control 10.2.0.4.0.
1 windows 2003 server with Oracle Databases 10.2.0.3.
1 windows 2003 server with Oracle Databases 10.2.0.3 - Data Guard setup
I have tried to stop (and delete) some scheduled fullbackup jobs from Grid Control.
The job has now status SUSPEND. I have tried to stop and delete the job, but I get the message:
All executions of the job were stopped successfully. Currently running steps will not be stopped.
I cannot delete the job and the job is not running.
I have run this procedure in order to delete the job:
DECLARE
jguid RAW;
BEGIN
SELECT job_id
INTO jguid
FROM mgmt_job
WHERE job_name = '<name of your job>'
AND job_owner = '<owner of your job>';
mgmt_job_engine.stop_all_executions_with_id(jguid,TRUE);
COMMIT;
END;
With no effect. The job is still shown in Grid Control with status SUSPEND.
I have restarted all servers and all the components in Grid Control, but the jobs will not disappear from Grid Control although they have been deleted.
I am struggling with this for about 2 days now.
I have search in Metalink and the internet, but I have not found anything that provides a solution.
Any help will be very much appreciated.

hi,
I have in the past used the following from metalink
Do you have a metalink account?
SET verify OFF
SET linesize 255
SET pagesize 128
SET trimout ON
SET trimspool ON
SPOOL jobdump.log
ALTER SESSION SET nls_date_format='MON-DD-YYYY hh:mi:ss pm';
COLUMN status format a15
COLUMN job_name FORMAT a64
COLUMN job_type FORMAT a32
COLUMN job_owner FORMAT a32
COLUMN job_status format 99
COLUMN target_type format a64
COLUMN frequency_code format a20
COLUMN interval format 99999999
VARIABLE JOBID VARCHAR2(64);
PROMPT *********************** JOB INFO ********************************
REM Get the job id
SET serveroutput on
BEGIN
SELECT job_id INTO :JOBID
FROM MGMT_JOB
WHERE job_name='&&jobName'
AND job_owner='&&jobOwner'
AND nested=0;
EXCEPTION
WHEN NO_DATA_FOUND THEN
BEGIN
DBMS_OUTPUT.put_line('JOB NOT FOUND, TRYING NAME ONLY');
SELECT job_id INTO :JOBID
FROM MGMT_JOB
WHERE job_name='&&jobName'
AND nested=0
AND ROWNUM=1;
EXCEPTION
WHEN NO_DATA_FOUND THEN
DBMS_OUTPUT.put_line('JOB NOT FOUND');
END;
END;
SELECT job_name, job_owner, job_type, system_job, job_status, target_type
FROM MGMT_JOB
WHERE job_id=HEXTORAW(:JOBID);
PROMPT *********************** JOB SCHEDULE ****************************
SELECT DECODE(frequency_code,
1, 'Once',
2, 'Interval',
3, 'Daily',
4, 'Day of Week',
5, 'Day of Month',
6, 'Day of Year', frequency_code) "FREQUENCY_CODE",
start_time, end_time, execution_hours, execution_minutes,
interval, months, days, timezone_info, timezone_target_index,
timezone_offset, timezone_region
FROM MGMT_JOB_SCHEDULE s, MGMT_JOB j
WHERE s.schedule_id=j.schedule_id
AND j.job_id=HEXTORAW(:JOBID);
PROMPT ********************** PARAMETERS ********************************
SELECT parameter_name,
decode(parameter_type,
0, 'Scalar',
1, 'Vector',
2, 'Large', parameter_type) "PARAMETER_TYPE",
scalar_value, vector_value
FROM MGMT_JOB_PARAMETER
WHERE job_id=HEXTORAW(:JOBID)
AND execution_id=HEXTORAW('0000000000000000')
ORDER BY parameter_name;
PROMPT ********************** TARGETS ********************************
SELECT target_name, target_type
FROM MGMT_JOB_TARGET jt, MGMT_TARGETS t
WHERE job_id=HEXTORAW(:JOBID)
AND execution_id=HEXTORAW('0000000000000000')
AND jt.target_guid=t.target_guid
ORDER BY target_type, target_name;
PROMPT ********************** FLAT TARGETS ********************************
SELECT target_name, target_type
FROM MGMT_JOB_FLAT_TARGETS jft, MGMT_TARGETS t
WHERE job_id=HEXTORAW(:JOBID)
AND jft.target_guid=t.target_guid
ORDER BY target_type, target_name;
PROMPT ************************ EXECUTIONS *******************************
SELECT execution_id,
DECODE(status,
1, 'SCHEDULED',
2, 'RUNNING',
3, 'FAILED INIT',
4, 'FAILED',
5, 'SUCCEEDED',
6, 'SUSPENDED',
7, 'AGENT DOWN',
8, 'STOPPED',
9, 'SUSPENDED/LOCK',
10, 'SUSPENDED/EVENT',
11, 'SUSPENDED/BLACKOUT',
12, 'STOP PENDING',
13, 'SUSPEND PENDING',
14, 'INACTIVE',
15, 'QUEUED',
16, 'FAILED/RETRIED',
17, 'WAITING',
18, 'SKIPPED', status) "STATUS",
scheduled_time, start_time, end_time
FROM MGMT_JOB_EXEC_SUMMARY e
WHERE job_id=HEXTORAW(:JOBID)
ORDER BY scheduled_time;
UNDEFINE jobName
UNDEFINE jobOwner
UNDEFINE JOBID
SPOOL OFFAlan

Groupwise 8 on Server 2008R2 SP1 Performance and Defrag Questions

I am running GroupWise 8.0.2 on an HP DL360G G7 with 24 GB of RAM and Dual Xeon X5650 processors under Server 2008 R2 Sp1. Post Office is using roughly 562 GB of an 819 GB Disk located on an HP p2000 g3 SAS connected enclosure comprised of 12 x 146 GB (15k rpm) SAS disks in a RAID 10 configuration. Typically never more than 700 users connected to the Post Office at any time. I am experiencing client slowness at regular intervals. Roughly every three weeks. The solution tends to be to restart the server and give it a fresh start, which holds for a while.
Concerns.
1. When the server restarts I see minimal memory utilization maybe 1.5 GB within half an hour the Used memory climbs a bit to about 2GB but free memory is quickly pre-allocated to standby memory. Within about 2 hours the Free memory is all consumed and I have about 20+ GB of Standby Memory. Running RAMMap indicates that the memory is being used by Metafile and Mapped File, Which tells me that the Post Offices files are being indexed and cached into RAM. Then after a couple weeks go by the amount of active RAM exceeds 8GB (still mostly Metafile, not so much Mapped File). And standby RAM still consumes the remaining RAM between Metafile and Mapped file, leaving no free memory. Typically once I reach about 8GB of memory actively used by the system (mostly Metafile), it's time for performance to drop for the clients.
Also, I'm seeing increases in Disk Queue Length for the Post Office Volume. Typically below 2 rarely as high as 5,
I suspect my best solution is to start a regular defrag process as my volume is now 29% defragmented [yeah NTFS :( ]
Question:
I am concerned a defrag could take as much if not longer than 10 hours. Which is too long for the agents to be down. So I was wondering if anyone had used DiskKeeper or alternative 3rd party Defrag utilities that might be able to defrag open files in the background, or if anyone had run defrag with the agents running to get the defragable files, then shut down the agents for a second pass which should be considerably shorter. Any advice that can be offered, or other suggestions for my described issue, would be greatly appreciated.
Thank You!

In article <[email protected]>, Matt Karwowski wrote:
> I am running GroupWise 8.0.2 on an HP DL360G G7 with 24 GB of RAM and Dual Xeon X5650
> processors under Server 2008 R2 Sp1. Post Office is using roughly 562 GB of an 819
> GB Disk ... I am experiencing client slowness at regular intervals. Roughly every
> three weeks. The solution tends to be to restart the server and give it a fresh start,
A) Updating to latest code may assist as this could also be a memory fragmentation type
issue or such that has been fixed
B) Perhaps even more RAM might help. How much space do the ofuser/*.db and ofmsg/*.db files
take up? The more mail flows, the more of those DB file content are held in memory. A few
versions ago, you tended to need as much RAM as the total of the DB files, and while those
days are gladly past, it is still a good number to watch out for.
C) Explore Caching mode for at least some of the users as that significantly reduces the
load on a server.
> I suspect my best solution is to start a regular defrag process as my volume is now
> 29% defragmented [yeah NTFS :( ]
Still way better than old FAT ;)
This is one of the reasons why GroupWise runs better on Linux with either EXT3 or NSS
volume types, and is why SLES is provided along with your GroupWise license.
As for running Defragments, even running it for just a few hours a week will gradually
help, especially if you can fit in one big burst near the beginning. So if you can automate
the process of shut down agents, run defrag for X hours and then shut it down, then restart
the agents, you may clear this all up. Just having the agents down an hour a week might
clear the memory usage issue for you.
I would be very hesitant to run any defrag on open database files of any email system
unless the defrag tool knew explicitly about that email system. But a smart defragger that
can keep the *.db files in their own (fastest) section of the drive and the rest off to the
side would go a long way to making the defragmentation process much more efficient.
I haven't directly run a GroupWise system on any flavor of Windows, since OS/2, so this is
more a combination of all my platform knowledge, but I hope it gets you closer to a smooth
running system. And if the other GW on Windows admins can pipe in, all the better.
Andy Konecny
KonecnyConsulting.ca in Toronto
Andy's Profile: http://forums.novell.com/member.php?userid=75037

Oracle 10.2.0.3 Databases - Restore to New Server

We will soon be replacing an old server with a new server and I need to restore all Oracle 10.2.0.3 databases to the new server
and ensure Oracle functionality/connectivity with all databases.
I am in the process of testing and need advice and hopefully a 'Step by Step' process I can follow.
I am a systems admin - not an Oracle db admin - so please explain details of any commands I may have to run.
I have already tested once:
ran command prompt: sqplus / as sysdba
SQL>shutdown immediate;
Stopped all Oracle services:
Oracle Database
Oracle Listener
Oracle DBConsole
Installed Oracle 10.2.0.3 on new server
Ensured that the SID and all other Oracle services information matched the old server during installation on the new server.
Stopped all Oracle services on the new server (used the steps above).
Backed-up and test restored to new server.
Enterprise Manager shows: Listener Up Database Instance Down Agent Down
Any help would be appreciated

It would be good to know your operation system. I hope both new and old match or similar.
You can use RMAN to dumplicate you DB so you can follow [Creating a Duplicate Database on a New Host. Doc ID: 388431.1|https://support.oracle.com/CSP/main/article?cmd=show&type=NOT&id=388431.1]
but a simple file system copy based manual duplication is also possible if new and old OS system is similar...

Agent Down

Similar Messages

Maybe you are looking for