Patch makes instance unavailable

After applying patch 8.1.5.0.2 I get an error when trying to connect to the database
ORA-01096 program version (8.1.5.0.2) incompatible with instance 8.1.5.0.0
Suggestions? After applying the patch I did not see any log file errors. I shut down the database prior to applying but perhaps it did not shutdown properly.

Similar Messages

Agent Connection to Instance Unavailable

Win2k3 64 bit
Oracle Database 10g Release 10.2.0.5.0 - 64bit Production
I brought up a new 10.2.0.5.0 db on Win2k3 64 bit server. Everything is working fine, at least it seems so. This server has been running for 2 months now and this issue kept coming up in the last 2 months.
Database Instance status is Open
Listener status is Up
General status is Up
Agent Connection to Instance is Unavailable.
Users SYSMAN and DBSNMP are OPEN.
This will be the 3rd time in the last 2 months if I rebuild EMCA again.
Is anyone else besides me experiencing this issue? Why is this happening and how do I fix it permanently?

1.Post the output of :
SQL> select global_name from global_name;
SQL> show parameter db_domain;
2.lsnrctl service
3.lsnrctl status
4.Did you change the hostname?
5.Is timezon of agent and console the same?
6.In the file $AGENT_HOME/sysman/config/emd.properties, make sure the following parameters are correct. Obtain the values from $OMS_HOME/sysman/config/emd.properties
1. REPOSITORY_URL
2. emdWalletSrcUrl
3. emdWalletDest (ensure that this destination is present on your management server).
7.If you're using Data Guard in 11g for primary/standby configuration, you may want to check the Data Guard configuration status. With this type of error, it's usually caused by the database parameter setting for log archive that differs with the Data Guard config for log transport and apply. For example, if the local archive log destination on the primary is pointing to the standby site instead of local archive directory (or local flash_recovery_area), The GC will generate this error and blank out the options for database instance in GC.
I am just pasting the above texts from different forum link when I just said "Agent Connection to Instance Unavailable" to google.
Regards
Girish Sharma

How to Patch Multiple Instances on same oracle home

Hi All
I Have 4 Instances running on server like below
DEVSTAG /opt/oracle/product/10.2.0
DEVUAT /opt/oracle/product/10.2.0
DEVREVT /opt/oracle/product/10.2.0
DEVREG /opt/oracle/product/11.2.0
i want to Apply the patch set 11.1.0.7.1 to DEVUAT & DEVREVT Instances , Can someone please help .. i dont want this patch set applied to DEVSTAG Instance .. how to exclude that Instance from getting patched ?
Thanks in advance .. hope my requirement is clear.

Which patch are you referring to ? The database patching tool (opatch) generally applies any database patch to an ORACLE_HOME. You will need to follow the instructions in the README of the patch. If you need to only patch one instance, then you will have to create a dedicated ORACLE_HOME for each database (i.e they will not share database executables). If you share ORACLE_HOME amongst multiple instances, then you will have to patch all of them simultaneously.
HTH
Srini

Database instance unavailable

Hi all,
I am getting an exception : Database instance unavailable when i want to execute following command
emca -config dbcontrol db
emca -config dbcontrol db
STARTED EMCA at May 24, 2013 3:24:28 PM
EM Configuration Assistant, Version 11.2.0.0.2 Production
Copyright (c) 2003, 2005, Oracle. All rights reserved.
Enter the following information:
Database SID: hms
Exception in thread "main" oracle.sysman.emcp.exception.DatabaseUnavailableException: Database instance unavailable.
        at oracle.sysman.emcp.DatabaseChecks.throwDBUnavailableException(DatabaseChecks.java:151)
        at oracle.sysman.emcp.DatabaseChecks.checkDbAvailabilityImpl(DatabaseChecks.java:144)
        at oracle.sysman.emcp.DatabaseChecks.checkDbAvailability(DatabaseChecks.java:163)
        at oracle.sysman.emcp.DatabaseChecks.getDbServiceName(DatabaseChecks.java:582)
        at oracle.sysman.emcp.EMConfigAssistant.performConfiguration(EMConfigAssistant.java:1267)
        at oracle.sysman.emcp.EMConfigAssistant.statusMain(EMConfigAssistant.java:573)
        at oracle.sysman.emcp.EMConfigAssistant.main(EMConfigAssistant.java:521)I already recreated repository twice.
Thanx

[root@iaf-db ~]# export oracle_sid=hms
[root@iaf-db ~]# echo $oracle_sid
hms
[root@iaf-db ~]# emca -config dbcontrol db
STARTED EMCA at Jun 5, 2013 9:43:05 AM
EM Configuration Assistant, Version 11.2.0.0.2 Production
Copyright (c) 2003, 2005, Oracle. All rights reserved.
Enter the following information:
Database SID: hms
Exception in thread "main" oracle.sysman.emcp.exception.DatabaseUnavailableException: Database instance unavailable.
        at oracle.sysman.emcp.DatabaseChecks.throwDBUnavailableException(DatabaseChecks.java:151)
        at oracle.sysman.emcp.DatabaseChecks.checkDbAvailabilityImpl(DatabaseChecks.java:144)
        at oracle.sysman.emcp.DatabaseChecks.checkDbAvailability(DatabaseChecks.java:163)
        at oracle.sysman.emcp.DatabaseChecks.getDbServiceName(DatabaseChecks.java:582)
        at oracle.sysman.emcp.EMConfigAssistant.performConfiguration(EMConfigAssistant.java:1267)
        at oracle.sysman.emcp.EMConfigAssistant.statusMain(EMConfigAssistant.java:573)
        at oracle.sysman.emcp.EMConfigAssistant.main(EMConfigAssistant.java:521)
EMCA LOG
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.ParamsManager setParam
CONFIG: Setting param: SID value: hms
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.DatabaseChecks getDbServiceName
CONFIG: No service name available. Will try to set db_unique_name.db_domain
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.ParamsManager getParam
CONFIG: No value was set for the parameter REP_DB_RUNNING_HOST.
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.ParamsManager getParam
CONFIG: No value was set for the parameter DB_UNIQUE_NAME.
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.ParamsManager getParam
CONFIG: No value was set for the parameter ORACLE_HOSTNAME.
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.ParamsManager getHost
CONFIG: Host having instnces running for database null : iaf-db
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.ParamsManager setParam
CONFIG: Setting param: REP_DB_RUNNING_HOST value: iaf-db
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.ParamsManager getParam
CONFIG: No value was set for the parameter SYSMAN_PWD.
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.ParamsManager getParam
CONFIG: No value was set for the parameter PORT.
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.ParamsManager getParam
CONFIG: No value was set for the parameter ORACLE_HOSTNAME.
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.util.ClusterUtil isCRSInstalled
CONFIG: isCRSInstalled: false
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.util.ClusterUtil getLocalNode
CONFIG: Cluster.isCluster: false. Skip call to getLocalNode
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.util.ClusterUtil getLocalNode
CONFIG: isLocalNodeDone: true localNode: null
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.ParamsManager getParam
CONFIG: No value was set for the parameter DB_UNIQUE_NAME.
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.ParamsManager getRepDbPort
CONFIG: Couldn't get listener port. File /app/oracle/product/11.2.0/dbhome_2/iaf-db_null/sysman/config/emoms.properties does not exists
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.ParamsManager getParam
CONFIG: No value was set for the parameter SERVICE_NAME.
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.util.GeneralUtil isLocalConnectionRequired
CONFIG: isLocalConnectionRequired: true
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.util.GeneralUtil initSQLEngine
CONFIG: isLocalConnectionRequired: true. Connecting to database instance locally.
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.util.GeneralUtil initSQLEngineLoacly
CONFIG: SQLEngine connecting with SID: hms, oracleHome: /app/oracle/product/11.2.0/dbhome_2, and user:
Jun 5, 2013 9:43:08 AM oracle.sysman.emcp.util.GeneralUtil initSQLEngineLoacly
CONFIG: ORA-01031: insufficient privileges
oracle.sysman.assistants.util.sqlEngine.SQLFatalErrorException: ORA-01031: insufficient privileges
     at oracle.sysman.assistants.util.sqlEngine.SQLEngine.executeImpl(SQLEngine.java:1650)
     at oracle.sysman.assistants.util.sqlEngine.SQLEngine.connect(SQLEngine.java:973)
     at oracle.sysman.emcp.util.GeneralUtil.initSQLEngineLoacly(GeneralUtil.java:445)
     at oracle.sysman.emcp.util.GeneralUtil.initSQLEngine(GeneralUtil.java:637)
     at oracle.sysman.emcp.DatabaseChecks.checkDbAvailabilityImpl(DatabaseChecks.java:118)
     at oracle.sysman.emcp.DatabaseChecks.checkDbAvailability(DatabaseChecks.java:163)
     at oracle.sysman.emcp.DatabaseChecks.getDbServiceName(DatabaseChecks.java:582)
     at oracle.sysman.emcp.EMConfigAssistant.performConfiguration(EMConfigAssistant.java:1267)
     at oracle.sysman.emcp.EMConfigAssistant.statusMain(EMConfigAssistant.java:573)
     at oracle.sysman.emcp.EMConfigAssistant.main(EMConfigAssistant.java:521)

Patching multiple instances from single unzipped patch

Hi,
I would like to know if it is possible to patch two Oracle Apps instances on two different machines from common shared patch location.
For eg: two instances -ABC on S1 and XYZ on S2 are to be patched. The patch is unzipped in a common,shared location - /stage/<patch_no.>
Can I apply the same patch in the common location /stage , to both ABC and XYZ AT THE SAME TIME?
Thanks,
Sandeep

You can patch from a common NFS mountpoint. If you have different applications users, ensure they are all in the same group, and the staged mountpoint has group write permissions.
This works best on Linux. On Solaris, almost 100% of the time when the patch is more than 1 GB in total size, NFS hangs on the checking file versioning stage of the copy portion. The only way to clear the hung NFS is to reboot the server that is being patched, then restart adpatch. So on Solaris, I normally copy to local filesystem whenever the patch exceeds 1 GB.

Inkscape with cairo patch - "make" makes an error

hi, i'm trying to build inkscape with the cairo patch from the inkscape-devel list.
with this, writing pdf should work very fine!
http://sourceforge.net/mailarchive/foru … m_id=36054
here is the patch - it was in the mail: http://bloink110.de/inkscape/patch.tgz
first i copy the files from the patch in the source folder from inkscape0.44 and than i start the patch
first strange is:
patch -p0 < patch.bin
patching file src/extension/extension.h
patching file src/extension/internal/Makefile_insert
patching file src/extension/internal/pdf-out.cpp
Hunk #5 succeeded at 143 (offset 7 lines).
patching file src/extension/init.cpp
patching file src/extension/internal/Makefile_insert
Reversed (or previously applied) patch detected! Assume -R? [n]
makefile_insert is patching 2time with same substance.
ok, but when is goes to "make", there came an error.
extension/libextension.a(init.o): In function `Inkscape::Extension::init()':
extension/init.cpp:111: undefined reference to `Inkscape::Extension::Internal::PrintPDFCairo::init()'
collect2: ld returned 1 exit status
make[2]: *** [inkscape] Error 1
make[2]: Leaving directory `/home/markuman/eigenePKGs/nochmal/inkscape-0.44/src'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/home/markuman/eigenePKGs/nochmal/inkscape-0.44'
make: *** [all] Error 2
i don't know why?! - and cairo 1.2 is installed too.
in inkscape-devel list and irc chanell people tell that the patch works fine - but why it don't work here?
greetings
markus

hi, i'm trying to build inkscape with the cairo patch from the inkscape-devel list.
with this, writing pdf should work very fine!
http://sourceforge.net/mailarchive/foru … m_id=36054
here is the patch - it was in the mail: http://bloink110.de/inkscape/patch.tgz
first i copy the files from the patch in the source folder from inkscape0.44 and than i start the patch
first strange is:
patch -p0 < patch.bin
patching file src/extension/extension.h
patching file src/extension/internal/Makefile_insert
patching file src/extension/internal/pdf-out.cpp
Hunk #5 succeeded at 143 (offset 7 lines).
patching file src/extension/init.cpp
patching file src/extension/internal/Makefile_insert
Reversed (or previously applied) patch detected! Assume -R? [n]
makefile_insert is patching 2time with same substance.
ok, but when is goes to "make", there came an error.
extension/libextension.a(init.o): In function `Inkscape::Extension::init()':
extension/init.cpp:111: undefined reference to `Inkscape::Extension::Internal::PrintPDFCairo::init()'
collect2: ld returned 1 exit status
make[2]: *** [inkscape] Error 1
make[2]: Leaving directory `/home/markuman/eigenePKGs/nochmal/inkscape-0.44/src'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/home/markuman/eigenePKGs/nochmal/inkscape-0.44'
make: *** [all] Error 2
i don't know why?! - and cairo 1.2 is installed too.
in inkscape-devel list and irc chanell people tell that the patch works fine - but why it don't work here?
greetings
markus

When can I make instance variables public?

I noticed that some of the in-built Java classes have public instance variables, for example in Rectangle the variables x, y, width and height are all directly accessible from outside. When is it okay to make public instance variables? Is it when they are primitive types? I thought maybe it is when they are final variables - but that can't be the case because you can directly modify the Rectangle's public instance variables.
I've always used getter and setter methods in my objects because I was told "making instance variables public is bad". But obviously this isn't necessarily the case (unless the guys at Java are bad programmers). I think directly accessing the instance variables makes for much neater and more readable code. Is there any convention or best-practise guidelines on when I can or can not make them public?
P.S. I also just "don't get" why you can't make them public if you're just going to provide a getter and setter method anyway... what's the difference? I suppose the theory is that the setter method should do some checking to make sure the input is valid. But what if the setter method just accepts -anything- of the appropriate type (accepts any Rectangle, for example)? What's the use of hiding the instance variables?
teach me.
Thanks. ;-)

For the most part, it is true that making instance member variables other than private is not a good idea. This concept, known in OO circles as "information hiding", typically allows the greatest flexibility in internal implementation of a class, since you can change the internal structure of the class without affecting the clients that use the class.
Let me add that Sun's programmers are a team, and teams have members of differing capabilities.
It is also true that there may be cases where exposing instance variables is unlikely to cause problems. Simple classes, or very stable classes that will not change are some examples, but not all.
Following a guideline blindly is not a good idea either. You need to evaluate each instance in which you decide you might want to deviate from "good programming practices". Still, best practices are best practices for a reason. It is best to understand them, and then decide if you can violate them.
For my own part, I don't think it is "neater" to expose such variables, and I have personally had to deal with situations where a change in a class's internal structure was extremely painful because the original developer chose to make all his instance variables public.
{?

Patching new instance.

DBA's
I just created a new 10gR2 instance on a Solaris box using DBCA. The ORACLE_HOME which I used was already patched with the July06CPU. Do I have to run the catcpu.sql script against the new instance?
Thanks,
Kirk

DBA's,
I found out that the sys object REGISTRY$HISTORY contains all the version info for the latest patch installed in the database. When I query the old instance that object had the correct CPU info in it - the new instance did not. This leads me to believe that I would have to run the catcpu.sql on the newly created instance. Is this a correct assumption?

Mplayer patch: make mplayer work with growing files

Mplayer currently doesn't seem to handle the growing files. They are usually used when you stream something from the web (e.g. youtube-dl). Now when mplayer reaches the offset which was the size at the open time it just quits with various errors. Here's a simple way to reproduce. Get/download a movie and do the following (let's name our movie b.mp4):
$ size=$((2*1024*1024))
$ head -c $size b.mp4 > a.mp4
Now start mplayer a.mp4. Tweak the size parameter to have around ~10 seconds of movie. Mplayer will quit when it reaches the 10th second. Now do this:
$ head -c $size b.mp4 > a.mp4
$ [start the movie in the background and wait until it begins to play]
$ tail -c +$((size+1)) b.mp4 >> a.mp4
Mplayer will still quit at the 10th second even though it could continue playing (VLC continues to play without a hitch).
Mplayer used to continue playing but it was because ffmpeg was buggy and stopped working after the following patch has been introduced: http://patches.libav.org/patch/14105/
I'm not sure how to fix this properly but this patch seems to fix this on my computer. There's a related bug to this and this patch seems to fix that too: http://bugzilla.mplayerhq.hu/show_bug.cgi?id=1821
I've sent this patch upstream, but I got no response, so I thought I post it here in case anybody else wants to try it (or try to send it upstream again):
Index: stream/cache2.c
===================================================================
--- stream/cache2.c (revision 34788)
+++ stream/cache2.c (working copy)
@@ -579,6 +579,15 @@
return (cv->max_filepos-cv->read_filepos)/(cv->buffer_size / 100);
+off_t cache_maxpos(stream_t *s)
+{
+ cache_vars_t *cv;
+ if (!s || !s->cache_data)
+ return -1;
+ cv = s->cache_data;
+ return cv->max_filepos - 1;
+}
+
int cache_stream_seek_long(stream_t *stream,off_t pos){
cache_vars_t* s;
off_t newpos;
Index: stream/cache2.h
===================================================================
--- stream/cache2.h (revision 34788)
+++ stream/cache2.h (working copy)
@@ -24,5 +24,7 @@
void cache_uninit(stream_t *s);
int cache_do_control(stream_t *stream, int cmd, void *arg);
int cache_fill_status(stream_t *s);
+/// Return the offset of the last byte in the cache.
+off_t cache_maxpos(stream_t *s);
#endif /* MPLAYER_CACHE2_H */
Index: libmpdemux/demux_lavf.c
===================================================================
--- libmpdemux/demux_lavf.c (revision 34788)
+++ libmpdemux/demux_lavf.c (working copy)
@@ -99,6 +99,8 @@
stream_t *stream = demuxer->stream;
int64_t current_pos;
mp_msg(MSGT_HEADER,MSGL_DBG2,"mp_seek(%p, %"PRId64", %d)\n", stream, pos, whence);
+ off_t maxpos = stream->end_pos = cache_maxpos(stream);
+ if (maxpos+1 > stream->end_pos) stream->end_pos = maxpos+1;
if(whence == SEEK_CUR)
pos +=stream_tell(stream);
else if(whence == SEEK_END && stream->end_pos > 0)

Thank you for this. I've been experiencing this problem and I find it very disruptive to my normal work-flow. For now, I've downgraded to mplayer version 34426-3, since I don't want to compile the whole of mplayer. Hopefully upstream can get this issue sorted out.

Patching Java instances with hotfixes using MOPZ - downgrade?

If hotfixes for a certain part of the J2EE engines are installed and MOPZ is used to get a new support package, it takes only the core patches without the hotfixes.
E. g. if BI_REPLAN SP14 + Hotfix 3 is installed MOPZ will pick BI_REPLAN SP15 Hotfix 0. This leads to a warning in JSMP that I´m "downgrading" an application.
As far as I understood those "hotfixes" they are like (example)
COMPONENT SP14 Hotfix 3 is equal
COMPONENT SP15 Hotfix 2 is equal
COMPONENT SP16 without hotfix
Since the application in the SMP does not enable me to manually pick the .scas with the latest hotfixes and I can´t manually download them since they are not in the necessary JSPM-.xml file what´s the procedure here? Can I safely assume that all applications can be "downgraded" without harming the system?
Why is MOPZ not picking up the latest ones?
Markus

> In your case, I think it is not recommended to downgrade. I think it is the case to find out the next support package available ( after the hot fix ).
Well - if would do like this we would still be on SP12 and not, as recommended, on SP15.
> If you need to download a fix, you still can do this going to SAP Service Marketplace, find the hot fix package and put it in your download basket manually. Then, in order to approve the download, you will need to go through MOPZ ( this time it will show you the current download basket with the fix already there and you will be able to download it ).
Yes - the problem is, I can't integrate that fix into the stack - and it's not always possible to wait until a stack fixes the problem "by default".
> In a last option, you can open an OSS message requesting the file for the hot fix. SAP will give you an URL where you can download it whithout passing through MOPZ.
MOPZ is not the problem, the problem is, that those hotfixes which are available in the normal download area are not taken from MOPZ. I'm not talking about hotfixes as attachment to notes, however, the problem is the same there.
If MOPZ is "the tool to manage" deltas then it should do appropriately and not just "half"
Markus

What Are The Condions that Make OpenCL Unavailable?

Up to 2 days ago I had ATI Catalyst 12.2 installed. I was able to check all the boxes in the Advanced Graphics Processor Settings dialog and they stayed set. Specifically, I had [ ] Use OpenCL set and Photoshop CS6 worked very well.
Because I need to know about the quality of newer drivers for my own product development, I upgraded my ATI driver suite to the current version - Catalyst 12.4.
Unfortunately, this caused new problems I hadn't had before - specifically a number of GPU-accelerated features didn't work, such as Oil Paint, which would try to start then just exit. And with Catalyst 12.4 the [ ] Use OpenCL box became grayed-out.
With my own testing done, I decided to completely uninstall the ATI Catalyst 12.4 suite, and drop back to Catalyst 12.2, which worked well. This was a simple matter of uninstalling everything AMD then rebooting (with the Windows driver), then installing Catalyst 12.2. That went smoothly.
However, now I still do not find the [ ] Use OpenCL box has become available again - it's grayed out:
I'm pretty sure there are no remnants from the 12.4 drivers left anywhere on my system, and everything from the 12.2 set installed successfully, so the question is this: Did Photoshop sense something about the 12.4 driver set, then set some kind of permanent "No OpenCL" flag in its own configuration settings?
Here's my system info with the 12.2 drivers installed. Note the "OpenCL: Unavailable" entry.
http://Noel.ProDigitalSoftware.com/ForumPosts/SystemInfo.txt
-Noel

Oops, I posted it in the wrong forum (in addition to misspelling "Conditions").
Pierre (or other forum Mod), could you please just delete this thread? I have reposted it over on the Photoshop General Discussion forum.
Thanks.
-Noel

What Are The Conditions that Make OpenCL Unavailable?

Up to 2 days ago I had ATI Catalyst 12.2 installed. I was able to check all the boxes in the Advanced Graphics Processor Settings dialog and they stayed set. Specifically, I had [ ] Use OpenCL set and Photoshop CS6 worked very well.
Because I need to know about the quality of newer drivers for my own product development, I upgraded my ATI driver suite to the current version - Catalyst 12.4.
Unfortunately, this caused new problems I hadn't had before - specifically a number of GPU-accelerated features didn't work, such as Oil Paint, which would appear to try to start then just exit a split second later. And with Catalyst 12.4 the [ ] Use OpenCL setting became unavailable (grayed-out).
With my own testing done, I decided to completely uninstall the ATI Catalyst 12.4 suite, and drop back to Catalyst 12.2.
This was a simple matter of uninstalling everything AMD then rebooting (with the Windows driver), then installing Catalyst 12.2. That went smoothly.
However, now I still do not find the [ ] Use OpenCL box has become available again - it's grayed out:
I'm pretty sure there are no remnants from the 12.4 drivers left on my system (though there's a possibility I could have missed something), and everything from the 12.2 set installed successfully, so the question is this:
Did Photoshop sense something about the 12.4 driver set, then set some kind of permanent "No OpenCL" flag in its own configuration settings?
Here's my system info with the 12.2 drivers installed. Note the "OpenCL: Unavailable" entry.
http://Noel.ProDigitalSoftware.com/ForumPosts/SystemInfo.txt
Note that all my GPU-accelerated features seem to work... Oil Paint works and is plenty fast, the various Blur tools work nice and fast, I can do 3D rendering (though there's one small glitch I've had since Photoshop CS6 was released where occasionally it becomes unresponsive for a few seconds at a time while doing 3D work). But it does finally respond.
-Noel

Thanks for the info.
I understand what OpenCL is about, and to some extent at least the code should be portable, since there are standards that the implementations follow.
But whether Adobe's code IS portable to all video cards, or whether all video drivers/cards implement the same standards, or even whether all video drivers/cards implement the standards correctly is always in question.
From what I can see whether Photoshop CS6 is willing to consider the OpenCL in the ATI/AMD Catalyst 12.2 package good enough to use comes and goes. The [ ] Use OpenCL box may be grayed out in some runs of Photoshop and not in others. I've personally set the option, had Liquify come up and run fine, and then later (in another run of Photoshop CS6) found [ ] Use OpenCL unchecked and grayed-out - and STILL had Liquify come up and run fine. I suspect this is because there are a number of different ways Photoshop is willing to try to do the work, and it's falling back to an alternate strategy.
In Catalyst 12.4 I would NEVER see the OpenCL setting available, and in fact Liquify wouldn't even come up.
I assume the support for what Photoshop needs is marginal in the recent ATI/AMD drivers, but it would be good to know a little more about why the OpenCL setting comes and goes like it does.
-Noel

Mutiple instance for messaging server

Can I set up mutiple instance of messaging server in one physical server ? The purpose is to separate domain to use difference resource, configuration. Now we use JES verion 6.

Can I set up mutiple instance of messaging server in
one physical server ? The purpose is to separate
domain to use difference resource, configuration. Now
we use JES verion 6.While I don't know if it supported(I think it wasn't but we needed to make it work while JES is still in transition about HA strategy), I did an installation of 3 JES2004Q2 messaging instances on 3 different hosts that were all in the same Veritas Cluster. With this configuration, we were able to have any messaging instance run on any host in the cluster (and could shift them so all 3 instances were running on a single host without problem, which we did several times)
Each host had a SAN partition mounted as
/usr/iplanet-$host
such as
/usr/iplanet-host1
/usr/iplanet-host2
these partitions were able to be moved between hosts as needed by the cluster software.
On each host, I made sure it had the SAN mount in place, then installed messaging as:
/usr/iplanet-$host/msg61
(instead of /opt/SUNWmsgsrv)
and when I ran /usr/iplanet-$host/msg61/sbin/configure
I said that the data directory should be:
/usr/iplanet-$host/msg61/msg-$host
I installed the admin server as:
/usr/iplanet-$host/admin5
On each of the messaging instances, I ran the ha_ipconfig script to lock down the IP address that the server processes bound to (and also had to lock down the ens daemon using a separate configutil variable, it still listened on the wildcard address).
When I was configuring the admin servers, I made each instance's run on a seperate port number so I didn't need to worry about locking down the IP. We also had to make a slightly customized version of the VCS agent that Sun supplies to work with the model we decided on.
There are still some issues, such as with performing patches. To patch an instance, you would need to make sure it is running on the host that it was originally installed on. If you only have a single host, I don't think the technique I used would work for you (unless you are able to use Solaris 10 beta and the zones functionality, and I am not sure if the JES packages play well with it yet)

Webkit unavailable. Failed to start UMP:(9) executable not found. bold 9790

i am unavailable to open inbox as well browser. message displayed on screen : shared webkit instance unavailable. failed to start UMP: (9) executable not found. as well it has start rebooting itself once till now. kindly help out me.
Thanks

jawediqbal wrote:
i am unavailable to open inbox as well browser. message displayed on screen : shared webkit instance unavailable. failed to start UMP: (9) executable not found. as well it has start rebooting itself once till now. kindly help out me.
Thanks
I suggest you do a clean Reload or Update of your Operating Sysytem. Go here for instructions: How To Reload Your Operating System
Goodluck!
Click if you want to Thank someone. If Problem is resolved, so that others can make use of it.

*.pls need to recompile after revised? How to make it active after revised?

Dear all,
as the subject, i revised the code of POXVDCKB.pls, does it need to recompile? How to make it active?
My environment is : ORALCE EBS R12.1.2 | AIX 6.1
Regards
Terry

as the subject, i revised the code of POXVDCKB.pls, does it need to recompile? How to make it active?This is a seeded file, any reason you had to edit it?
To run the pls file, you have to connect to SQL*Plus and execute the file (just like executing any SQL script).
You may also need to run the script to recreate the package specification.
Please note that you will lose all the changes if you patch the instance with a patch that delivers the same file.
Thanks,
Hussein

Patch makes instance unavailable

Similar Messages

Maybe you are looking for