I'm a huge fan of sym-link to the "production" version of any package.
That way, you can have older, or newer, releases available simultaneously.
(see end)
The usual magic prefix for IBM is /usr/lpp.
So maybe point /usr/lpp/javaprod at the current release.
You could then set JAVAHOME in
On 21/01/2016 16:50, Rob Schramm wrote:
You are being short sighted. Why would I spend additional time and energy
updating a myriad of JAVA_HOME statements, scripts, JCL when I can simply
standardize the whole thing. Sure you can always override JAVA_HOME... but
that is the exception.. Not the
On Thu, 21 Jan 2016 03:14:44 +, Rob Schramm wrote:
>About the /etc/profile...
>Best to set JAVA_HOME to a symlink. Less work for an upgrade. The users
>can always override such things.. Having a standard symlink(s) that can be
>used gives a nice amount of control and consistency.
>
Can you
On 21/01/2016 15:00, Jack J. Woehr wrote:
Andrew Rowley wrote:
In answer to the original question, having SMP/E use the JAVA_HOME
environment variable would be the clean way.
http://www-01.ibm.com/support/knowledgecenter/SSLTBW_2.1.0/com.ibm.zos.v2r1.gim3000/java.htm?lang=en
Isn't that
Andrew Rowley wrote:
Isn't that javahome from a different location? Not the JAVA_HOME environment
variable?
Yeah, but it's set the same way JAVA_HOME is, pointing to the root of the
installation. IBM always has to be kinky,
and beside, they probably wanted to make sure that if the user has
Paul Gilmartin wrote:
Can you cite an authority for "JAVA_HOME"?
https://docs.oracle.com/cd/E19182-01/820-7851/inst_cli_jdk_javahome_t/
http://askubuntu.com/questions/175514/how-to-set-java-home-for-java
$JAVA_HOME is a convention, but it's a strong one recognized by umpteen
applications,
About the /etc/profile...
Best to set JAVA_HOME to a symlink. Less work for an upgrade. The users
can always override such things.. Having a standard symlink(s) that can be
used gives a nice amount of control and consistency.
Rob Schramm
On Wed, Jan 20, 2016, 5:30 PM Farley, Peter x23353 <
Andrew Rowley wrote:
In answer to the original question, having SMP/E use the JAVA_HOME environment
variable would be the clean way.
http://www-01.ibm.com/support/knowledgecenter/SSLTBW_2.1.0/com.ibm.zos.v2r1.gim3000/java.htm?lang=en
--
Jack J. Woehr # Science is more than a body of
Follow on story about 18 year old server
http://www.theregister.co.uk/2016/01/20/486_fleet_still_in_production/
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with
Andrew,
You are being short sighted. Why would I spend additional time and energy
updating a myriad of JAVA_HOME statements, scripts, JCL when I can simply
standardize the whole thing. Sure you can always override JAVA_HOME... but
that is the exception.. Not the rule.
Rob Schramm
On Thu, Jan
On 21/01/2016 14:14, Rob Schramm wrote:
About the /etc/profile...
Best to set JAVA_HOME to a symlink. Less work for an upgrade. The users
can always override such things.. Having a standard symlink(s) that can be
used gives a nice amount of control and consistency.
Except that JAVA_HOME is
Ed Gould wrote:
>Follow on story about 18 year old server
According to below article, it seemed that 18 year server is just a mere infant
...
>http://www.theregister.co.uk/2016/01/20/486_fleet_still_in_production/
Hehehe. Thanks Ed. I really enjoyed that follow-up article.
The day will come
I have come up against a problem I can seem to solve. I am using HMC
Version 2.13.0.
If I am local to the HMC, everything works as expected. The same is true
for all functions
EXCEPT for Single Object Operations when I am coming thru a Browser.
I select SOO and select the "Target Object", then
Our LPARs are being hosted on a single CEC which also might include
other client LPARs. All I'm seeing in operlog is that the STP timing
network for this CEC doesn't have backup or arbiter servers assigned.
Mark Jacobs
Vernooij, CP (ITOPT1) - KLM wrote:
Do you really get a stand-alone CEC or
Yea, that was my thought too. Just wanted to ask around. Thanks.
Mark Jacobs
R.S. wrote:
W dniu 2016-01-20 o 15:52, Mark Jacobs - Listserv pisze:
Our LPARs are being hosted on a single CEC which also might include
other client LPARs. All I'm seeing in operlog is that the STP timing
network
Sorry to relate the bad news of Dennis Malarky
I am sorry to pass on this sad news. Dennis was diagnosed with
cancer in November and didn't have a chance to have chemo, he entered
hospice on 12/8 and died on 12/12 at 12:30 am. He was interred in
the Riverside National Cemetery on
I do not, but wish that I did. I am in a JES3 environment right now and have
been in JES2 environments my entire 30+ year career. I will be closely
monitoring this thread. I have been reviewing a Redbook that deals with just
this issue. I have not gotten very far into it yet.
Thanks,
Rob
You can see this if you are coming though a NAT based router. You would
have to set up special port forwarding rules on the router.
Tony Thigpen
Steve Beaver wrote on 01/20/2016 09:50 AM:
I have come up against a problem I can seem to solve. I am using HMC
Version 2.13.0.
If I am local to
W dniu 2016-01-20 o 15:52, Mark Jacobs - Listserv pisze:
Our LPARs are being hosted on a single CEC which also might include
other client LPARs. All I'm seeing in operlog is that the STP timing
network for this CEC doesn't have backup or arbiter servers assigned.
Well, it is single CEC, so
Can you use this as justification to get to the two Nationwide sessions planned
at SHARE in San Antonio?
Cheryl
-Original Message-
From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf
Of Jackson, Robin W. Contractor
Sent: Wednesday, January 20, 2016 11:10 AM
Things to check:
1. java
2. certificates
Ad 1. The easiest method is to use HMC with downlevel java version.
Other means: Control Panel - Java - Security at High level (actually the
lowest possible), also fill Exception Site List with https://ip.add.re.ss
Ad 2. You have to allow self-signed
http://www.findagrave.com/cgi-bin/fg.cgi?page=gr=156689949
On Wed, Jan 20, 2016 at 12:50 PM, Ed Gould wrote:
> Sorry to relate the bad news of Dennis Malarky
>
> I am sorry to pass on this sad news. Dennis was diagnosed with cancer in
> November and didn't have a
I have noticed that it also depends upon HOW you got to various ISPF
panels. For example IN most cases when you enter 3.4 from the primary
panel enter END, you go back to where you were before 3.4 On the other
hand if you enter 3;4 from the primary panel (assuming ";" is your command
Really! At last!
Many years ago I submitted an SR; Security; against tcsetattr() where
I suspected the problem lay. (Before Ported Tools; I found breakage
with my own code.) IBM whitewashed the problem in "stty -echo"
but left tcsetattr() broken. I suppose their attitude was, "Don't do
that!"
Hi Gil,
I'm not sure I understand exactly what you are referring to?
Kirk Wolf
Dovetailed Technologies
http://dovetail.com
On Wed, Jan 20, 2016 at 5:14 PM, Paul Gilmartin <
000433f07816-dmarc-requ...@listserv.ua.edu> wrote:
> Really! At last!
>
> Many years ago I submitted an SR;
I vote symlink as well.
Rob Schramm
On Wed, Jan 20, 2016, 4:45 PM John McKown
wrote:
> I like the SYMLINK idea. I do that myself, "by hand".
>
> On Wed, Jan 20, 2016 at 3:34 PM, Paul Gilmartin <
> 000433f07816-dmarc-requ...@listserv.ua.edu> wrote:
>
> >
Doesn't z/OS Unix have the equivalent of /etc/profile? And doesn't Java
installation create an entry or invoked script in /etc/profile to set the
JAVAHOME environment variable as every other *ix system does? Shouldn't that
be the preferred method for compatibility with every other *ix Java
Dammit! The path to Java changes with any z/OS release and/or
any Java release. I need continually to add to my PATH variable
to keep up. And there's nothing an ISV can supply in JCL samples
for SMP/E's SMPJHOME; the example in the SMP/E Reference is
woefully outdated.
This makes as little
I like the SYMLINK idea. I do that myself, "by hand".
On Wed, Jan 20, 2016 at 3:34 PM, Paul Gilmartin <
000433f07816-dmarc-requ...@listserv.ua.edu> wrote:
> Dammit! The path to Java changes with any z/OS release and/or
> any Java release. I need continually to add to my PATH variable
> to
On 2016-01-20 15:28, Farley, Peter x23353 wrote:
> Doesn't z/OS Unix have the equivalent of /etc/profile?
>
It has.
> And doesn't Java installation create an entry or invoked script in
> /etc/profile
>to set the JAVAHOME environment variable
>
it doesn't
> as every other *ix system does?
>
I
W dniu 2016-01-20 o 03:11, Dale R. Smith pisze:
On Tue, 19 Jan 2016 12:33:44 -0800, Charles Mills wrote:
"(c) International Business Machines Corporation 1974, 1976, 1981"
35 to 42 years old!
Charles
The last supported release of the OS/VS COBOL Compiler and Library was
It's being configured that way by our outsourcer, we no longer have
control over the configuration. I agree with you, but wanted to see if
there's any additional risk to our environment in this configuration.
Mark Jacobs
Vernooij, CP (ITOPT1) - KLM wrote:
The main reason of STP is to
Do you really get a stand-alone CEC or are your LPARs placed on one of the
outsourcers machines?
If so, they will probably have STP and provide a reliable time.
And of course Gil is right: if not, you can configure STP to a timeserver to
keep your time accurate.
Kees.
-Original
On Wed, 20 Jan 2016 07:08:18 +, Vernooij, CP (ITOPT1) - KLM wrote:
>The main reason of STP is to coordinate time between the CECs. When you have
>only one, there is nothing to coordinate anymore.
>
Some reference is necessary to steer the TOD clock in order to prevent
secular drift. What
34 matches
Mail list logo