cancel
Showing results for 
Search instead for 
Did you mean: 

APS service start up issue with BI 4.0

Former Member
0 Kudos

We are running BI 4.0 Sp2, Patch 15 in our DEV environment on Windows. All of a sudden this afternoon we starting having issue with the BO services in one of the 2 clustered boxes. First Tomcat wouldn't start at all, failing at every try. After trying a server reboot, all other services are coming up except for the Adaptive Processing Server. I enabled trace on APS and after sifting through a lot of info, found this:

M  [Thr 3600] ***LOG Q0I=> NiPGetHostByName: '$SMDAgentHost' not found: getaddrinfo [ninti.c 895]

M  [Thr 3600] *** ERROR => Main agent failed to be initilized due to incorrect host setting! [ncsmtdatasen 22]

M  [Thr 3600] *** ERROR => Main agent failed to be initilized due to incorrect host setting! [ncsmtdatasen 22]

M  [Thr 3600] NCS trace timer is disabled since the interval is set to 0

M  [Thr 3600] NCS data timer is disabled since the interval is set to 0

M  [Thr 3600] NCS library version 2.3.8 (unicode) loaded

M  [Thr 3600] NCS_ProcInit API invoked

Any idea what might be going on? the APS services just don't want to come back up! Also, we have been unable to uploaded universes/objects to the CMS since this started happening even though the CMS services and I/OFRS seem to be up and running.

Thanks!

Accepted Solutions (1)

Accepted Solutions (1)

hi Sarang,

when i checked your error post it says the following

incorrect host setting!

can you please check the host file settings in your windows box if you are using a windows server.

I have faced a similar issue with BO 3.2 and resolved by correcting the host file.

please do check there and if you have any entry for BO servers or clusters try disabling all of them and try again.



Henry_Banks
Product and Topic Expert
Product and Topic Expert
0 Kudos

Sarang hasn't been on this post for 6+ months..

@David -  I'm kinda surprised that you're company is running BI 4.0 SP2 Patch 5 .

That was the minimum version that was ever shipped/generally available to the market ( back in mid 2011) that was effectively the base release, as everything before it was pre-release/ramp-up software.

It's not clear how much RAM is available to your BIPlatform, nor if you have 'split' the APS (i.e. create a dedicated APS just for the MDAS use-case),  nor is it clear how much -Xmx?G you have 'bumped up' in the command line.

Here's how to split APS :

here's a BIP sizing guide

the problem with your version is that we have since (several times) upgraded the verious JVMs to later and more stable versions.. so it's very hard to make a meaningful assesment, because all these issues were potentially fixed at the time (almost 3 years ago)

Thanks & regards,

H

Answers (1)

Answers (1)

Henry_Banks
Product and Topic Expert
Product and Topic Expert
0 Kudos

Hi,

the sap host agent and ncs instrumentation would be used by the Monitoring application in CMC.

i had this problem after patching.  a number of settings had reverted: Platform Search was indexing again, Monitoring Application was enabled, Auditing was back on etc

you should disable the 3 things above, and stop unecessary processing servers. then your platform will stabilise.

this problem may be indicative of a shortage of memory resource.

Regards,

H

Former Member
0 Kudos

I have this issue too,but don't know why it is,and how to deal it,could you help me?

---------------------------------------------------

trc file: "jobserver_XMTMEDBBO.AdaptiveJobServer_WebIJavaSchedulingService_CHILD0_ncs.trc", trc level: 1, release: "720"

---------------------------------------------------

M [Thr 9816] Thu Dec 27 11:08:05 2012

M  [Thr 9816] ***LOG Q0I=> NiPGetHostByName: '$SMDAgentHost' not found: getaddrinfo [ninti.c 895]

M  [Thr 9816] *** ERROR => Main agent failed to be initilized due to incorrect host setting! [ncsmtdatasen 22]

M  [Thr 9816] *** ERROR => Main agent failed to be initilized due to incorrect host setting! [ncsmtdatasen 22]

M  [Thr 9816] NCS trace timer is disabled since the interval is set to 0

M  [Thr 9816] NCS data timer is disabled since the interval is set to 0

M  [Thr 9816] NCS library version 2.3.8 (unicode) loaded

M  [Thr 9816] NCS_ProcInit API invoked

0 Kudos

hi henry

thanks for the information regarding the error i am running BI 4.0 in windows server getting the same error message

---------------------------------------------------

trc file: "cms_DEV_BOBJSIA.CentralManagementServer_ncs.trc", trc level: 1, release: "720"

---------------------------------------------------

M [Thr 2520] Mon Feb 11 12:01:55 2013

M  [Thr 2520] ***LOG Q0I=> NiPGetHostByName: '$SMDAgentHost' not found: getaddrinfo [ninti.c 895]

M  [Thr 2520] *** ERROR => Main agent failed to be initilized due to incorrect host setting! [ncsmtdatasen 22]

M  [Thr 2520] *** ERROR => Main agent failed to be initilized due to incorrect host setting! [ncsmtdatasen 22]

M  [Thr 2520] NCS trace timer is disabled since the interval is set to 0

M  [Thr 2520] NCS data timer is disabled since the interval is set to 0

M  [Thr 2520] NCS library version 2.3.8 (unicode) loaded

M  [Thr 2520] NCS_ProcInit API invoked

M  [Thr 2520] NCS_ProcExit API started

M  [Thr 2520] Timer threads are successfully finished

M  [Thr 2520] Data sending threads are successfully finished

M  [Thr 2520] NCS_ProcExit API invoked

Is there any way you could help me with fixing it

thanks

denis_konovalov
Active Contributor
0 Kudos

do not look at the *trc file, you will not find usefull info there, enable HIGH trace level in CMC for the server troubling you and then look at its *.glf log.

0 Kudos

Thanks Denis for your swift response actually i have followed your advice and set

bo_trace.ini file

active = true

important = xs

and restarted the SIA node and Tomcat in the following order

1. SIA

2. Tomcat

Looked in to the eventvwr for errors. I am getting the TNS:listener does not currently know of service requested in connect descriptor error.

But if i do a tnsping from command prompt the same data base responds with OK message.

Thanks for any help and suggestions

denis_konovalov
Active Contributor
0 Kudos

by editing bo_trace.ini you just enabled tracing for the whole suite.

If you're having a problem with just one server, you need revert changes to bo_trace.ini, and enable HIGH trace for that specific server in the CMC.

Then look at the trace (*.glf) generated for that specific server.

tnsping and the error you describe have nothing to do with each other.

0 Kudos

hi Denis,

the original issue i had was user not able to login to CMC and receiving server not available error.

I tried to fix the issue through checking all firewall and rules setting in server , but even with fire wall turned off i got the same message . (Server *****:6400 not found or server may be down (FWM 01003)).

Then today i started looking in SCN for other available solutions and switched on the trace with the help of server admin team. Like you mentioned me i have switched on the trace for the particular server alone and also i checked in the *.glf file not in the trace files.

TraceLog message 33

|d56dadd2-0131-53e4-98af-7ee2770d291f|2013 02 11 14:21:47:546|+0000|Information| |==| | |cms_DEV_BOBJSIA.CentralManagementServer| 2820|2824|| ||||||||||||||||||||||SIServerController:run: Fail to initialize or uninitialize.  Reason is The root server reported an error Initialization Failure. (Reason: SAP BusinessObjects BI platform CMS: Unable to connect to the CMS system database "*****". Reason: ORA-12514: TNS:listener does not currently know of service requested in connect descriptor

SAP BusinessObjects BI platform CMS: Unable to connect to the CMS system database "****". Reason: ORA-12514: TNS:listener does not currently know of service requested in connect descriptor

CDatabase::Open failure.

This is the message am getting from the glf file(cms_DEV_BOBJSIA.CentralManagementServer_trace.000001).

When i do a TNS ping from command prompt i am able to get response from the data base

Is this some thing you can help me with

thanks

karthick

denis_konovalov
Active Contributor
0 Kudos

then you have hijacked the original poster thread with an unrelated issue.

Please open your own thread.

The vast majority of the time, this error results from an incorrectly specified connect descriptor in the tnsnames.ora file

0 Kudos

thanks denis for your help till now in this issue have opened a new question for the issue.

Former Member
0 Kudos

I realize this is an old thread but it wasn't flagged as officially answered and this is one that I need the answer to.  We started experiencing this issue out of the blue about two weeks ago when I tried to open an Analysis document that I had created a few days before.  The following MDAS error came up:

  • "The requested operation could not be completed. Verify that the MDAS Service is running and that the database client software has been installed on the machine hosting the MDAS Service, and try again."

When I went into the CMC I noticed that the APS was showing Failed.  After a failed Start I decided to check the log/trace files.  I see the following in the trace file:

M  [Thr 10388] ***LOG Q0I=> NiPGetHostByName: '$SMDAgentHost' not found: getaddrinfo [ninti.c 895]

M  [Thr 10388] *** ERROR => Main agent failed to be initilized due to incorrect host setting! [ncsmtdatasen 22]

M  [Thr 10388] *** ERROR => Main agent failed to be initilized due to incorrect host setting! [ncsmtdatasen 22]

M  [Thr 10388] NCS trace timer is disabled since the interval is set to 0

M [Thr 10388] Tue Oct 29 11:10:17 2013

M  [Thr 10388] NCS data timer is disabled since the interval is set to 0

M  [Thr 10388] NCS library version 2.3.8 (unicode) loaded

M  [Thr 10388] NCS_ProcInit API invoked

M [Thr 10988] Tue Oct 29 11:10:17 2013

M  [Thr 10988] NCS_ProcExit API started

M  [Thr 10988] Timer threads are successfully finished

M  [Thr 10988] Data sending threads are successfully finished

M  [Thr 10988] NCS_ProcExit API invoked

From what I gather, and correct me if I'm wrong, this is somewhat of a generic error.  In looking at the GLF log file I see the following Java error each time I attempt to start the service:

java.lang.UnsatisfiedLinkError: com.sap.tracelog.NativeMethods$Version_4_0.config_create_BAB68345(IIZZLjava/lang/String;Ljava/lang/String;IIIIIILjava/lang/String;Ljava/lang/String;ZIIII)J

at com.sap.tracelog.NativeMethods$Version_4_0.config_create_BAB68345(Native Method)

at com.sap.tracelog.NativeMethods$Version_4_0.access$4800(NativeMethods.java:47)

at com.sap.tracelog.NativeMethods.config_create(NativeMethods.java:568)

at com.sap.tracelog.Native$Marshaler.makeNativeConfig(Native.java:247)

at com.sap.tracelog.NativeMethods.state_create(NativeMethods.java:584)

at com.sap.tracelog.Native$Marshaler.makeNativeState(Native.java:274)

at com.sap.tracelog.NativeMethods.deviceAdmin_state(NativeMethods.java:378)

at com.sap.tracelog.JNIHelpers.pushMainDeviceConfigToCxx(JNIHelpers.java:26)

at com.businessobjects.framework.servers.platform.bootstrap.Main.setupTracelog(Main.java:796)

at com.businessobjects.framework.servers.platform.bootstrap.Main.initializeLogging(Main.java:234)

at com.businessobjects.framework.servers.platform.bootstrap.Main.initialize(Main.java:440)

at com.sap.js.startup.Engine.<init>(Engine.java:69)

at com.sap.js.startup.Main.main(Main.java:66)

I tried what you suggested, changing the settings on the three items, but still no luck getting the APS to start.  I also tried bumping up the memory in the  'Command Line Parameters' section of the APS service but still no luck.  And finally, I tried both cloning and creating a new instance of the service, all with the same results.  The SIA has been restarted several times in addition to the server being rebooted.

We're currently running BI 4.0 SP2 Patch 5 and Java Version 6 Update 24.  Any help/advice would be greatly appreciated!

Thanks,

Dave

Former Member
0 Kudos

Hi Dave,

Did you ever resolve this issue? I am receiving the same error upon restart of my DEV environment servers (now on 4.1 SP2 but I believe started a while ago).

Thank you,

Jen

Former Member
0 Kudos

Hi Jen,

No, this issue is still outstanding.  I'm waiting for a ticket to be logged by our support team (our environment is not configured properly in the Support Portal).  I will let you know when I hear anything, but you may have figured it out by then (in which case you will have to let me know ).

Thanks

Dave

Former Member
0 Kudos

Hi Dave,

Thank you for your response.  Please do let me know when the ticket is logged so that I can track the outcome as well - and/or let me know any outcomes you find along the way.  I will do the same.

Best,

Jen