OS X from a Linux admin's perspective

I was tasked with getting an rsync daemon running on a lion box for backups from linux machines.
Problems:
1.) The rsync version shipped with lion is buggy and outdated.
2.) Updating it requires a workaround solution using macports (which requires xcode.)
3.) Downloading xcode requires that I create an account on the "App Store" and find the box, hook up a monitor/keyboard/mouse, and enter my credit card information.
4.) The xcode install made me hunt down and kill some "itunes helper" process before it would continue.
5.) The installer just locked up for no apparent reason.
6.) It didn't lock up cleanly; I had to kill a few processes to be able to get it restarted.
7.) Command line invocation of xcode programs still resulted in an error that no Java runtime was installed (I confirmend that a java runtime was in fact installed.)
This is not a troll. I like apple stuff, but they seem to be totally failing as servers.

"Did you have a question?" <-- funny... useless, but funny.
Anyway, valid concerns Jarl, and while I can't answer to all of it I can share some notes.
The App Store issue is bad, both for servers and anyone who cares about things like that being forced on you. A lot of can't use it even if we wanted to. No real idea what to tell you there, but as far as I can tell you can move your XCode install to another machine without license issues. In fact, a lot of apps seem OK like that. That's how I get them on my machines which don't have App Store access.
Regarding outdated utilities: this is true and there is no default package manager for OSX either. However, you can use something like macports. I think even pkgsrc works for OSX.
However, my recommendation would be Homebrew. I find it much simpler than things like macports, but still very powerful. Its based on Ruby and its very easy to write your own package installers if its catalog does not have the program or the version that you want.
Xcode4: You can run xcodebuild from a command line, but I have not found a way yet to configure everything about a project without the GUI, and some of the application testing tools need a GUI. There is no technical reason for this, its just a glaring omission in XCode for now. I used Apple's feedback page to tell them what I expect from XCode and suggest that for any developer who wants this fixed. Still once you have a project set up, you might be able to then stage the build to other machines for building with xcodebuild on the command line.
I don't have tremendous time to deal with this, so I just use make files most of the time on OSX. If XCode really behaves badly, then I just install the GNU toolchain from a package system so I can get the code built.
One ugly solution you could try is to run VNC on the headless servers and export the display to a desktop machine, so you can use XCode and other tools that way from a single machine with a head. At least it keeps you from having to go put a head on your servers.
Homebrew: http://mxcl.github.com/homebrew/

Similar Messages

How to open a RPD, from a linux folder? OBIEE 11g

Hi,
How to open a RPD, from a linux folder? Is there a tool, or any command line command to open the RPD?
Thanks for your time and help.

Hi,
You cannot open an rpd on a Linux manchine.
You have to install the client (Admin tool) on windows and open the rpd remotely in online mode
Thanks
Deva

DB2 Migration from Windows - Linux supported with Backup/Restore ?

Hi folks,
we have to do a DB2 V9.1 migration from Windows -> Linux. Both servers are Intel based. Is it official supported by SAP to do this without SAP migration tools? E.g. with a backup/restore or redirected restore? I heard about it, but i'm not sure.
Thanks a lot
Jochen
Edited by: Jochen Raab on Mar 1, 2010 5:40 PM

Hi Jochen,
Please have a look at the DB2 docu.
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/topic/com.ibm.db2.luw.admin.ha.doc/doc/c0005960.html
It clearly stats that Linux and Windows are not compatible.
DB2® database systems support some backup and restore operations between different operating systems and hardware platforms.
The supported platforms for DB2 backup and restore operations can be grouped into one of three families:
Big-endian Linux® and UNIX®
Little-endian Linux and UNIX
Windows®
A database backup from one platform family can only be restored on any system within the same platform family. For Windows operating systems, you can restore a database created on DB2 Universal Database (UDB) V8 on a DB2 Version 9 database system. For Linux and UNIX operating systems, as long as the endianness (big endian or little endian) of the backup and restore platforms is the same, you can restore backups that were produced on DB2 UDB V8 on DB2 Version 9.
So you need to do an eport/import.
Regards,
Joachim

Connect to MSSQL from OWB Linux.

Hi,
This is almost an embarrasing question to ask because I should be able to find this information but Alas! such is life.
So here goes, I want to be able to connect from my OWB 10gR2 on Linux to Microsoft SQL Server on some windows box. I went into MY_PROJECT->Databases->Non-Oracle->SQL Server and hit create.
I got details for a location that I filled as follows -
User Name: username
Password: password
Host: Host\Instance
Port: 1433
Service Name: Instance
Use Global Name: unchecked
Schema: username
When I test this, I get -
ORA-28545: error diagnosed by Net8 when connecting to an agent
Unable to retrieve text of NETWORK/NCR message 65535
ORA-02063: preceding 2 lines from OWB_1087
Am I missing something? Do I need to do something else? I saw this posting -
Connect to MS SQL server from OWB on Linux platform
but it seems dated.
I don't know if there are easier ways to connect from OWB Linux to MSSQL Windows in the newer versions. Any help is greatly appreciated.
Thanks.

I recommend watching this viewlet: http://www.oracle.com/technology/products/warehouse/htdocs/sql_server_source_all/sql_server_source_all_viewlet_swf.html
It shows how to setup connections to SQL Server. However, there is one major change. The viewlet shows how to do this on Windows and you need to do this on Linux. The only step that changes is step 1: setting up the ODBC driver. You will need to find a Linux ODBC driver and driver manager for SQL Server and get that setup. After that you can use the viewlet to setup the Heterogeneous Services.
Bryan

What is the best client for ARD from ubuntu linux?

I attempt to connect to my mac mini running osx 10.7 from ubuntu linux using the default vncviewer application and the tightvnc client and I get to the login display but am unable to click on any of the users nor log in. I was successful from my windows7 laptop, but not from my windows xp HTPC--ultravnc gets to the login prompt, showed me the password field, but didn't allow me to change focus to that field, however, this isn't reproducable since I get a different behavior each time I try to remote in.

Back up all data. Update all third-party software to the latest version and remove any you don't need. Download the Yosemite installer from the App Store. Run it.

Passing a FILENAME from a LINUX SHELL Script to an SQLPLUS Script

I written a LINUX Shell Script to receive a FILENAME. This FILENAME is the name of the file that I want to SPOOL into.
So, below are two items.
The first item is the LINUX Shell Script that has the FILENAME
The last item is the SQLPLUS Script that is suppose to receive the name
The LINUX Shell Script:
#! /bin/sh
LOGFILE="$(date +"%Y-%m-%d %H:%M") Oracle REPORT LOG.TXT"
FILENAME="$(date +"%Y-%m-%d") PROD Unfinished Spot Bids.CSV"
echo "File Name for the spool is:" "$FILENAME"
#spool "$FILENAME"
#sqlplus /nolog @run_test_query.sql
#sqlplus -S @Unfinished_Spot_Bids_Run_V1.sql >$LOGFILE
sqlplus -S /nolog <<EOF >$LOGFILE
/* I want to pass to the SQL file below */
/* The value that is contained in the the variable $FILENAME */
/* How do you do that? */
@@Unfinished_Spot_Bids_V1.sql
spool
set echo on
EOF
RV=$?
#if {$RV -ne 0}
#exit $RV
#else
echo "The return code fronm the sqlplus was:" "$RV"
#fi
Here is the SQLPLUS Script
CONNECT glog_read/parker@OTMPROD
set linesize 32000 trimspool on pagesize 3000
set echo off
SET TERMOUT OFF
set heading off
/* I want to parameterize the name of the file below */
/* This file is executed from inside a LINUX Shell Script */
/* That LINUX SHELL script is passed a file name */
/* I want that filename used in the "SPOOL" statement below */
/* So, how do I do that? */
spool PROD_Unfinished.CSV
prompt Shipment Update Date|Shipment Number|No of Shipments|Auctions per Shipment|No. of Carriers Notified|No of Bidders|Duration of Auction|Elapse time for first bid|Elapse time for last bid|Elapse time of bidding|Lead Time of Auction to P/U|High Bid- Low Bid Diff|
set feedback off
SELECT
to_char(c.update_date, 'FMMonth, YYYY')||'|'|| b.SHIPMENT_GID
||'|'|| count (distinct (b.SHIPMENT_GID))
||'|'|| count(distinct(a.i_transaction_no))
||'|'|| count(*)
||'|'|| sum (case when a.responding_gl_user_gid is null then 0 else 1 end)
||'|'|| min (numtodsinterval((b.expected_response - b.insert_date),'day'))
||'|'|| min (numtodsinterval((a.update_date - a.transaction_time),'day'))
||'|'|| max (numtodsinterval((a.update_date - a.transaction_time),'day'))
||'|'|| numtodsinterval((max(a.update_date) - min(a.update_date)),'day')
||'|'|| numtodsinterval((max(b.shipment_time) - max(c.insert_date)),'day')
||'|'|| (max(a.bid_amount) - min(a.bid_amount))
from tender_collab_servprov a, tender_collaboration b, shipment c
where a.i_transaction_no = b.i_transaction_no
and b.shipment_gid not in (select sqa.shipment_gid from tender_collaboration sqa, tender_collab_servprov sqb where sqa.i_transaction_no = sqb.i_transaction_no and sqb.acceptance_code = 'A')
and a.i_transaction_no in (select i_transaction_no from tender_collaboration where tender_type = 'Spot Bid')
and b.shipment_gid = c.shipment_gid
AND c.update_date >= trunc(last_day(to_date((to_char(add_months(SYSDATE,-2), 'MM-DD-YYYY')),'MM-DD-YYYY')) + 1)
group by c.update_date, b.SHIPMENT_GID
order by c.update_date, b.shipment_GID;
spool off

ParadicePGMR wrote:
First, I was getting an error on the SPOOL ${FILENAME}. Also, my file did have spaces and I did replace it with "_". It still did not work on the SPOOL Command. So, with or without the _, I never got the SPOOL command to work!
It seems that the FILENAME value from the LINUX script was not being passed into the SQLPLUS procedure.
How do you pass LINUX variables into the SQLPLUS procedure>
Not sure if I had the right usage => ${FILENAME}above is correct syntax

What do I need to open a connection from the Linux box to the Oracle server

Hi all,
I want to access oracle database from java application running on Linux.
The database is installed on a Window 2000 machine. What do I need to install before I can open a connection from the Linux box to the Oracle server on Windows 2000 ?
Do I need Oracle Client ?
Thanks,
Quoi

Hi Quoi,
http://myjdbc.tripod.com/basic/jdbcurl.html
Talks abt how to write a jdbc url and the jar + config required to connect to db. Also has a sample program to connect.
Regards
Elango.

DHCP does not load as a service from the Server Admin

Dear friends,
I have faced the following problem.
The task is to do internet sharing among other services on mac mini server.
Started with the clean install of OS X Lion, upgraded it to Server and then installed all updates.
USB Ethernet (en2) is looking into ISP with static IP, Ethernet (en0) is LAN.
Since I need specific address scope for the LAN I have actived DHCP from the Server admin and configured it to provide
192.168.8.11 to 192.168.8.200 as an adress scope, mask 255.255.0.0 on the interface en0.
Everything was working just fine for a couple of minutes and the LAN clients got their IP addresses from the DHCP scope. In a very short time Server Admin popped up and error message:
Try to refresh the view (aserver.local./Server). (kReceivedUnknownError)
Niether grey nor green bubble apears near the DHCP on the service list.
I keep recieving this error for all other services until I manage to disable dhcp on the services tab of server settings. After all I cant make it run again. I tried to restart dhcp from the terminal but it does not help. "sudo serveradmin fullstatus dhcp shows nothing".
Do you have any ideas how to make it work?

So you're not using NAT between the two? Even though you're using 255.255.0.0 so both network can talk (basically you have a class B network setup using Class C ip addresses), your network masks for each should be something like 255.255.255.xxx depending on how many subnets you need. NAT or Network Address Translation should be running and that will provide the interconnectivity between the two.
Your DHCP server should be set up only with 255.255.255.0 for the DHCP service since you are only serving IP addresses to the one subnet 192.168.8.x. Your starting address in DHCP setup is 192.168.8.11 ending address 192.168.8.200 and the netmask should be 255.255.255.0.

How to run ZAC CDP from a Linux Bundle

Hi all,
I am trying to work out how to run "ZAC CDP" from a Linux bundle. We have 330 Satelites that I need to force the replication of the TFTP folder too in one staggered process.
Is there any reason why this would not work? would a script do it and anyone have a sample?
cheers,
Trev.

I would imagine it's possible.
I'm assuming something (run as root user?)
#!/bin/bash
zac cdp
(you may have to put the absolute path in for the location of "zac")
I'm not a linux guru by any means, just that if you can script this via a cron job, I'm sure it can be done via ZCM Bundle.

What is the webconference tool of choice for demos from a linux box?

Since conference.oracle.com and strtc only work on IE, what are folks doing when they want to do a webconference from a linux box? Are people installing IE (I see it is possible but sounds a bit hacky), or is there another oracle conference tool or are people connecting from a PC to the linux box and running conference from there (is that possible?).
Thanks for any idea you could give me.

http://www.oracle.com/technetwork/middleware/beehive/downloads/index.html
http://forums.oracle.com/forums/category.jspa?categoryID=180

How do I access BSD file tree from within Linux?

I just installed PC-BSD 8.0 RC on an external USB HDD. I went with the PC-BSD default partition layout, i.e. one primary partition for the entire PC-BSD slice, and 4 "partitions" within that slice, for /, swap, /usr and /var.
In order to access the PC-BSD file tree from within Linux, I mount it as follows:
mount -t ufs -o ufstype=ufs2,ro /dev/sdb1 /mnt/bsd
But after executing this command I can only access the PC-BSD / , not /usr and /var , presumably because they reside on two BSD "partitions" (subslices) different from / . Is there any way I can mount these (read-only) in Linux, too, or would I have to reinstall PC-BSD and opt for laying out the slice differently, i.e. specifying only / and swap in sdb1?
Last edited by RobF (2010-02-13 11:55:39)

Look at /dev/sdb* there should be more than just sdb1. You would need to mount those to see which is which. For ufs write support the default kernel needs to be recompiled with experimental enabled.

Need Help Connecting To My Windows Server Machine From My Linux Machine (What Program To Use)

My name is Daniel, and I was wondering what program I could use to connect to my Windows Server R2 2012 Computer from my linux laptop. My laptop has the newest version of linux mint mate. I have tried nomachine with no success (but I could have been using
it wrong I do not really know how). So if anyone could help me out that would be great. Thank You.
P.S: Sorry for the weird Forum Category there is not really a category for this.

ok so i think that worked but i ran into another problem; what password should i use? i changed my sql server 2008 username/password and it didn't work. i also tried my windows 7 username/password and it didn't work also.
EDIT: managed to get it working, all good! thanks a lot.
Edited by: 1008695 on May 30, 2013 7:09 AM

Gracefully exit from a Linux kill

Is there a way to gracefully exit from a Linux kill command?
Given the following code:
public static void main(String[] args) throws IOException
    boolean bExit = false;
    while (!bExit)
      try
        Thread.sleep(1000L);
      catch (InterruptedException ex)
        bExit = true;
    System.out.println("Program exiting gracefully.");
}Is there a way for the Linux kill command to make the message "Program exiting gracefully." appear?

Certainly not for kill -9.
But you can look at Runtime.addShutdownHook().The problem I have with Runtime.addShutdownHook() is (from the API Javadoc):
"Shutdown hooks should also finish their work quickly. When a program invokes exit the expectation is that the virtual machine will promptly shut down and exit. When the virtual machine is terminated due to user logoff or system shutdown the underlying operating system may only allow a fixed amount of time in which to shut down and exit. It is therefore inadvisable to attempt any user interaction or to perform a long-running computation in a shutdown hook."
I just was wondering if there was a built-in hook for the kill command like the hook for "kill -3" that calls the Thread.interrupt() method on the main thread of execution.
My actual code is more like:
public static void main(String[] args) throws IOException
    boolean bExit = false;
    SubThreadManager1.init();
    SubThreadManager2.init();
    while (!bExit)
      try
        Thread.sleep(1000);
      catch (InterruptedException ex)
        bExit = true;
    // After these shutdown methods are called it could take up
    // to 30 seconds before they reach the stopped state.
    SubThreadManager1.shutdown();
    SubThreadManager2.shutdown();
    while (!SubThreadManager1.isStopped() || !SubThreadManager2.isStopped())
      try
        Thread.sleep(500L);
      catch (InterruptedException ex)
    System.out.println("Program exiting gracefully.");
}I was just simplifiying the test case to get the point across.

Upgrading wordpress from the wp admin panel. How?

Hi, I have major problems getting wordpress running properly with the LAMP settings found in the wiki. Thing is, the wordpress blog runs fine, but fails when updateing, installing or upgrading from the wordpress admin panel.
I've tried chown http:http, rataxes:rataxes, root:root and even a recursive chmod 777 /srv without succeding to update via the panel. Wordpress just says: "An unknown error occurred" when installing plugins. When upgrading I get wordpress's ftp update form in which I can't figure out what to enter.
I know some of you think it's a wordpress problem, it may partly be so, but the thing is it's only happening on my newly installed Arch machine. Any clues?

Have a look at PHP's error log (should be located at /var/log/user.log). My guess is that you need to set "allow_url_fopen" to "On" in /etc/php/php.ini.

Can I cloning the ebs from SUSE Linux 9 to Oracle Linux 5?

My EBS version is 11.5.10.2
Can I cloning the ebs from SUSE Linux 9 to Oracle Linux 5?

Hi,
My EBS version is 11.5.10.2
Can I cloning the ebs from SUSE Linux 9 to Oracle Linux 5?I belive you can not, you need to migrate your system. Please see:
Migrating to Linux with Oracle Applications Release 11i [ID 238276.1]
Debugging Platform Migration Issues in Oracle Applications 11i [ID 567703.1]
9i Export/Import Process for Oracle Applications Release 11i [ID 230627.1]
10g Release 2 Export/Import Process for Oracle Applications Release 11i [ID 362205.1]
Oracle Applications Installation Update Notes, Release 11i (11.5.10.2) [ID 316806.1]
You can also open SR and confirm it wiht oracle support
Regard
Helios

OS X from a Linux admin's perspective

Similar Messages

Maybe you are looking for