Symlinks treated differently in mavericks stacks

in mountain lion I used symlinks to reference most of the folders in my home directory to a hard drive connected to my time capsule. this worked as expexcted when I placed a stack on my dock: in grid veiw, a symlinked folder would open withn the stack. Since I have upgraded to mavericks symlinked folders open in a new finder window. is there some new way that mavericks stacks handles symlinks? can I change it back? is a herd link the way to go now? can I make hard links to directories across volumes???

OK, Installed Mavericks 10.9.2 and the behaviour is a bit different now.
It would appear that Apple has gone to their old ways and now when you put an "alias" in your home folders it will follow them AS symlinks.

Similar Messages

Vendor confirmation via EDI treated differently by MRP than Manually enterd

I have a strange situation with MRP.
I have a schedule agreement with the following information.
Date          Delivery sch            vendor confirmation
05/03/07          100                         100
05/10/07          100                         100
05/17/07          100                         100
05/24/07          100                         100
05/31/07          100                         100
06/07/07          100                         100
06/14/07          100                         100
Then I increase my independent requirments on 05/10/07 to 200.
If I have input my vendor confirmation manually via ME38, then MRP will add an additional Schedule line on 05/10/07 as I expected.   However if the vendor confirmations were sent in via idocs/edi then the schedule line is not added until the end of the confirmations.
Does anyone know why MRP would treat these differently based on if the confirmation were manual versus EDI?

It should not be treated saperately by MRP if it is posted manually or via EDI
i can think one thing that
do you have diff confirmaiton category for EDI and diff for manually than it can happen
so just make sure that all the confirmation category which you use is check for MRP relevant or not in
SPRO_MM_Purchasing-conformation-Set Up Confirmation Control-select your EDI confirmation contro key and click on confirmation sequence
and check here that the conf cat is check for MRP or not
if not than check than you will see that MEP will pickup that.

Cisco 3750 different models in stacked enviroment?

Will the c3750e-universalk9-mz.150-2.SE6.bin IOS on a Cisco Catalyst 3750X-48PF-E Switch and
the c3750-ipservicesk9-mz.150-2.SE6.bin IOS on a Cisco Catalyst 3750G-48PS Switch be able work in stacked mode?
For the 3750G switch there is no universal image. Im assuming that having a universal on the X series and ipservices on the G series would equate to each other for the different switch models. Is this correct?

Disclaimer
The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.
Liability Disclaimer
In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.
Posting
Again, "universal" image's feature set depends on license. As you have a 3750X IP Service license, effectively your universal image is an IP Services image.
Also again, you can mix IP Base and IP Services in the stack, but in your case, if you load an IP Services images on the 3750G, your two stack members will both be IP Services.
Yes, if you load the same IOS version on both switches, and they have IP Base or IP Services, they should work together. (You can also mix different IOS versions, but StackPlus versions need to match - which is best accomplished by using same IOS version.)
BTW, although can mix StackWise and StackWisePlus 3750s, the former will cause loss of some Plus features.

IOT without overflow treats differently char/varchar2

Hi,
test case with 8k block size:
Connected to:
Oracle Database 11g Enterprise Edition Release 11.2.0.1.0 - 64bit Production
With the Partitioning, Automatic Storage Management and Data Mining options
SQL>
     create table test1 (x number primary key, y char(2000)) ORGANIZATION INDEX;
Table created.
SQL>
     create table test2 (x number primary key, y varchar2(2000)) ORGANIZATION INDEX;
create table test2 (x number primary key, y varchar2(2000)) ORGANIZATION INDEX
ERROR at line 1:
ORA-01429: Index-Organized Table: no data segment to store overflow row-pieces
SQL>
     create table test2 (x number primary key, y varchar2(1000)) ORGANIZATION INDEX;
create table test2 (x number primary key, y varchar2(1000)) ORGANIZATION INDEX
ERROR at line 1:
ORA-01429: Index-Organized Table: no data segment to store overflow row-pieces
SQL>
     create table test2 (x number primary key, y varchar2(500)) ORGANIZATION INDEX;
Table created.why does it fail with varchar2 but not with char?
Thanks,
Joaquin Gonzalez
Edited by: user4070490 on 10-feb-2011 23:41
Maybe it is something different with characterset used for char and varchar:
     select * from V$NLS_PARAMETERS;
PARAMETER               VALUE
NLS_LANGUAGE            AMERICAN
NLS_TERRITORY           SPAIN
NLS_CURRENCY            €
NLS_ISO_CURRENCY        SPAIN
NLS_NUMERIC_CHARACTERS ,.
NLS_CALENDAR            GREGORIAN
NLS_DATE_FORMAT         dd/mm/yyyy hh24:mi:ss
NLS_DATE_LANGUAGE       AMERICAN
NLS_CHARACTERSET        AL32UTF8
NLS_SORT                BINARY
NLS_TIME_FORMAT         HH24:MI:SSXFF
NLS_TIMESTAMP_FORMAT    DD/MM/YYYY HH24:MI:SS.FF3
NLS_TIME_TZ_FORMAT      HH24:MI:SSXFF TZR
NLS_TIMESTAMP_TZ_FORMAT DD/MM/RR HH24:MI:SSXFF TZR
NLS_DUAL_CURRENCY       €
NLS_NCHAR_CHARACTERSET AL16UTF16
NLS_COMP                BINARY
NLS_LENGTH_SEMANTICS    CHAR
NLS_NCHAR_CONV_EXCP     FALSEddl after creation is:
CREATE TABLE "EXPLOTACION"."TEST1"
(    "X" NUMBER,
       "Y" CHAR(2000 CHAR),
        PRIMARY KEY ("X") ENABLE
) ORGANIZATION INDEX NOCOMPRESS PCTFREE 10 INITRANS 2 MAXTRANS 255 NOLOGGING
TABLESPACE "EXPLOTACION"
PCTTHRESHOLD 50;
CREATE TABLE "EXPLOTACION"."TEST2"
(    "X" NUMBER,
       "Y" VARCHAR2(500 CHAR),
        PRIMARY KEY ("X") ENABLE
) ORGANIZATION INDEX NOCOMPRESS PCTFREE 10 INITRANS 2 MAXTRANS 255 NOLOGGING
TABLESPACE "EXPLOTACION"
PCTTHRESHOLD 50;

its about size of colunm not datatype
SQL>/
create table test4 (x number primary key, y char(4000)) ORGANIZATION INDEX;
ERROR at line 1:
ORA-00910: specified length too long for its datatype
Elapsed: 00:00:00.56
SQL>create table test4 (x number primary key, y char(400)) ORGANIZATION INDEX;
Table created.
Elapsed: 00:00:00.28
SQL>

System refresh with different abap patch stacks

Hi gurus
I have a scenario that as of my ECC 6.0 PRD system has ABAP stack patch level at SP13. where my QAS system at stack level 17.
Now i have to refresh the QAS system with the PRD data.
My query is, is it possible to copy the PRD system to QAS( previously i did it successfully while both PRD & QAS was in the same stack 13, currenty i upgrade the QAS's SP stack to 17).
let me know whether is it possible to take a system copy from PRD?.... or any alternates..
environment is ECC 6.0 O/S: WIN2003 Server DB:MSSQL 2005
Thanks in advance.
Balaji nampally

Yes, it is possible to take a system copy from PRD.
But patch level is stored in the database, and so it will be copied.
Your QAS system will be at patch level 13 afterwards, like PRD.
If you want patch level 17 for QAS, you may either wait till PRD will be at 17, or you may repeat patching in QAS.
hope this helps

Are plain text files treated differently in SP 2010 than MOSS 2007?

We just moved to SP 2010 a week ago. A user just contacted us asking why the files she accesses are not displaying.
When she accessed these plain text files in MOSS 2007, they displayed within IE as a full page of text.
When she accesses these now in SP 2010, she is prompted to download the file.
These files do not end in .txt - they are names such as 2013.09.20 or whatever.
Is there a way to configure 2010 to display, without any interpretation, plain text files within the browser?
Thank you.

Hi ,
According to your description, my understanding is that you could not access the txt files with browser after migrading to SharePoint 2010..
Whether this issue occurred if you upload a new txt file to a library.
In my testing, everything worked well. The Open Documents in Client Applications by Default feature was acitve, and in the Library Settings->Advanced settings, ‘Use the server default(Open in the client application)’ was selected. When I clicked the file
name, it promted a dialog like the below, I selected ‘Read-Only’, and clicked OK, the file could open with browser.
Please go to IE->Tools->Internet Options->Programs->Manage add-ons, enable all add-ons related to SharePoint, compare the result.
I hope this helps.
Thanks,
Wendy
Wendy Li
TechNet Community Support

Autosuggest: XML and JSON datasets treated differently

Not sure if I found a bug or not, but I found something kinda
odd with Autosuggest and XML data vs. JSON data.
Check this url:
http://www.christopherchin.com/sandbox/testSpry.html
The first input box is XML data the second input is JSON.
When you type Christopher in both fields, you get a list of
people named christopher.
Now type a space and the first letter of the last name of one
of the Chris'.
XML works. JSON doesn't.
Thoughts?
Thanks!
Chris

Look at your JSON output, it has 2 spaces between the first
name and last name for some of your entries, but only in your JSON
output, not your XML output.
If you fix your JSON ouput so it matches the XML output, I'll
bet it works.
--== Kin ==--

Version Automatically on Top of Stack?

I may be dreaming, but I seem to recall in the early days of Aperture 3 (obviously not long ago), that whenever I created a version, by cropping, adjusting, etc, that new version was automatically placed on top of the stack.
That was very useful since when I edit a photo, I consider the edited version as my new number 1.
Now, however, the new versions seem to be placed at the end of the stack, requiring me to move them to the top.
Anyone have any comments on whether or how new versions can be set to automatically be placed on the top of a stack?
Thanks.

Gary wrote:
Gary -- Thanks for the numbers. Numbers are good to have.
PS I've also been using Stacks for grouping bracketed (and panoramic) sets. But I've also been using them for grouping adjusted versions. It doesn't seem to be quite working out - for me - so I might have to revise that.
My experience (limited, but aware) has been that, for each of the many grouping or binning functionalities in Aperture, I am always best served by choosing one and only one purpose. Your use of Stacks illustrates the problem: are Stacks for
. making picks among burst or bracketed shots?
. grouping panoramas and HDR singles (etc.)?
. grouping development versions of an image?
Stacks provide a sophisticated binning functionality (group, rank, collapse/expand) that is applicable to several tasks. Unfortunately, there is one and only one type of Stack, and no (easy) way to distinguish them. (One can easily imagine three kinds of Stacks, one each for the grouping needs listed above.) Am I right in assuming that this is the problem you imply with "+It doesn't seem to be quite working out+"?
For me, helping to winnow my photographs is the most important function I can assign to Stacks. I stack very similar shots, select the best of the Stack and promote it to the "Pick", and then go through the Project (I use one Project for each shoot) with Stacks collapsed and select the best among the Picks and the non-stacked shots.
(My "flower bed" is, from subsoil to bloom:
. Rejects
. All non-rejects
. Stack Picks and non-rejects
. Best of Project ("Selects", which are Project-specific)
. Selected for printing, export, publication, sale, etc. (Selected Selects -- which I move to Albums)
That leaves several things to address: how to handle bracketed shots, panorama shots, HDR shots, focus stacking shots, and, not least, Versions.
Shots bracketed in order to increase the odds of a proper exposure are easy: they get treated just like other Stacks. Stack 'em, pick the best, move on.
The other groupings I deal with in an "extra-computational" way: I remember that this stack is for a panorama (for instance). Then when I create the panorama I import it and put it in the Stack with the shots which comprise it, and make it the Pick. The "extra-computational" part is, clearly, a kludge, but a livable-with one.
Version-stacking muddies this tremendously (especially when the version is already in what I might call a "Picking Stack"). *I actually think that Aperture should add a specific and visually identified method of Version-stacking separate from the existing Stacks.* Every image whether inside or outside a user-defined Stack should have it's own Version stack. (Correct me if I'm wrong, but all Version of an image must be in the same Project, no?)
If I care to Pick two Versions in a Stack (either a Picking Stack or a Versions Stack), I split the Stack.
What is your Stacks workflow, and how might you revise it?
I'm also very curious about other's use of color labels. I find them queer, due to the fact that any image can have +one and only one+ color label. There is no set of useful identifiers I can think of which has 8 non-overlapping bins (I assume I am missing something terribly obvious). I tried using colors for different sets of identifiers, but failed at each because there were always images to which I wanted to assign two colored labels. I've ended up using the colors to assign steps along the progression of image development (Develop!, Developing (some adjustments applied but not done), Problem, Developed, Redevelop!, Redeveloped, Abandon, Not yet flagged).
This works well for me, because I am able to maintain separate identifier sets for development progression and image rating (I use the built-in stars). I can have good images not yet developed, developed images down-rated, images waiting to be rated or developed, and so on. What do you do?
Need I apologize for the ramble? I really want to set up a workable system that I then won't have to worry about for a couple of years and tens of thousand of images. Thanks for reading and any replies you might make.

CAS fails when subtracting negative numbers from stack

The CAS fails to subtract negative numbers. Here is how to reproduce such errors:1) generate a negative number from any operation1 - 22) subtract any number from this previous result2.1) enter any value12.2) press the subtract key, select the previous value from the stack, and copy it- (touch the previous "-1" which results from "1 - 2") (touch "copy")2.3) press ENTER Other variation:cos(-π)*ENTER*, results in -11 --1*ENTER* The issue is negative numbers are being preceded by a subtraction sign instead of a negative sign, and both are treated differently by the software. The bug does not occur in home mode.

Hello, Maké, Thank you! My report is that of a bug, not a device misuse. Particularly, a(nother) CAS bug. HP Prime's CAS is substituting binary subtraction signs for negative signs when printing results. Entering -1 in Home yedls -1, while in CAS one gets −1. See: the lengh of the dash symbol is different. Kind regards.

Pairing bluetooth presenter doesn't work on Mavericks anymore. Did work on SL.

I have a bluetooth remote presenter. It has the usual keys (next, previous, volume, play/pause, start default music player) and it also has a small touchpad / trackpad with left and right buttons, scrolling buttons.
The remote is great and it served me well under Windows and OSX 10.6.8 Snow Leopard. Now since I bought a new machine with Maverics, I can't pair it. The bluetooth utility looks different on Mavericks. Certain things missing, such as making Bluetooth discoverable or not in the way of simplification of things (Thanks Apple, but this wasn't big issue) or oversimplification of things, where there seems to be no access to Bluetooth device details and no way to add device without pairing code.
The remote is recognized as mouse with keyboard, which makes things worse, there is not only no code on the device, even if there is one (or if it would except 0000, 0000000, or similar), I have no way to enter it.
Unfortunately Maverics made things simple but useless for some bluetooth devices. It's unnecessary.
Seems like I will be stuck with Snow Leopard on one of my machines forever.
When I look in console on SL, here is what I see:
2014-06-28 8:23:11.268 AM System Preferences[7658]: [establishKernelConnection] Received an error from IOServiceOpen() - 0x10000003. NULLing out io_service_t.
There is an entry like this every time I try to pair.
The Bluetooth manager window says --
"Pairing failed"
When I go to Options, it shows the field to enter code. No matter what I enter there or if I leave empty, I won't be able to pair.
Otherwise BT works fine on Maverics, I have Apple keyboard, Logitech kayboard, Apple and Logitech trackpad and audio device connected, no issue.
The Remote presenter works also fine and I tried to re-pair it with other non-Mavericks machines and works fine too.

Hello,
Today I was trying to pair my MacBook Pro with a Microsoft Wireless Mouse 8000, which is a bluetooth mouse with no pairing code. I had used this mouse previously with Snow Leopard with (almost) no problems, but initially I couldn't find a way to pair this mouse in Mavericks, seemingly because the mouse lacks a pairing code and I must always enter one with this version of OS X.
Luckily I found a solution that worked for me. I'm going to post it here in case it helps polocanada or somebody else with a similar problem. In my case it involves using a BootCamp partition with Windows (Win7 in my case) with the Apple BootCamp drivers installed and started as a virtual machine using Parallels, but it may be possible to achieve the same result without a BootCamp partition (check the end of the post for more information)
These are the steps that need to be carried out:
1) Start the BootCamp Windows install using a virtual machine and connect the "Apple USB Bluetooth Host Controller" device to the virtual machine.
If BootCamp drivers were previously installed the Bluetooth device should be seen by Windows without problems.
2) Pair the mouse with the Windows install. This pairs the mouse with the Mac's bluetooth controller, just in the wrong OS.
3) Find out the mouse's Bluetooth ID. For Windows 7, right click the Bluetooth icon located in the task bar and click in Show Bluetooth devices. In the window that opens, right click in your mouse's icon and click Properties. Open the Bluetooth tab and copy somewhere the contents of the Unique Identifier field (six groups of two digits separated by colons). My Windows 7 is in Spanish so the texts may not be exactly the same as posted here in English.
4) Add the mouse's ID to the Mac. Open a terminal window in OSX and enter the following text: (substitute aa-bb-cc-dd-ee-ff with your mouse's Unique Identifier, changing the colons with hyphens)
sudo defaults write /Library/Preferences/com.apple.Bluetooth HIDDevices -array-add "aa-bb-cc-dd-ee-ff"
And enter your account's password when asked.
5) Disconnect the "Apple USB Bluetooth Host Controller" device from the virtual machine. The mouse should "just work" in OS X at this point.
Credit goes to this website. This is where I found the terminal command posted above.
In case you don't have a BootCamp partition with a Windows install or are not able to start the Windows install using a virtual machine, the link in the line above mentions a way to do the process with just terminal commands in the Manual Pairing section, but you still need a way to find out the mouse's Bluetooth ID.
Hope this helps!

Windows clients unable to access File Sharing since upgrading to Mavericks Server

This summer, I have been tasked to upgrade all the Macs on campus to Mavericks. Including upgrading to Server 3.1.2. Previously, we had an Xserve that ran 10.6.8 Server and two Mac mini servers that ran 10.8 Servers. We have a new Mac mini to replace the Xserve and, initially, everything was pretty smooth. I built the server from scratch, with the intention of using the 10.6.8 server as reference, only adapting to Profile Manager, instead of using Workgroup Manager.
The only problem I had when I started was that I couldn't get Active Directory to create home folders on my Mac Server "NetUsers" volume when adding a Home Folder path. After playing around with it, I completely hosed the Server and decided to start over. Now, after a clean install, I can't get any Windows users to connect to the share on the Mac Server. It declares: "\\MacServer is not accessible. You might not have permission to use this network resource. Contact the administrator of this server to find out if you have access permissions. Account restrictions are preventing this user from signing in. For example: blank passwords aren't allowed, sign-in times are limited, or a policy restriction has been enforced." What's more (that I may not have noticed two weeks ago or it just started, I don't know), the 10.8 server that I upgraded to 10.9 also is not accessible and gives the same error. However, I can access all these file shares from Macs using either smb or afp.
I'm suspicious that this issue is related to SMB1 vs. SMB2 that was implemented differently in Mavericks.
Please, any help on this would be a life-saver. School starts too soon and I need this running on day one!!!

Had the same issue and spent 4 solid days tracking it down (looking in all the wrong places.)
For us, it turned out to be a problem in DHCP services configuration.
For the subnet under the WINS tab, the NetBIOS Scope had a '1' inserted into the field. Apple's tag when hovering the mouse over the field indicated a domain be insterted.
This prevented the Windows XP machines from successfully registering with SAMBA. Once the field was blanked out and saved, the (one machine) needing to be joined found the PDC and successfully bound to the directory.
Blank out the NetBIOS Scope field in DHCP (assuming a simple LAN environment) and you may be in good shape.

Many of my System preferences won't load in mavericks?

Since updating to 10.9 many of my system preferences won´t load. These are:
- Security: You can’t open Security & Privacy preferences because it doesn’t work on an Intel-based Mac.
- Notifications: Could not load Notifications preference pane.
- Printers & Scanners: Could not load Printers & Scanners preference pane.
- iCloud: Could not load iCloud preference pane.
- Internet accounts: Could not load Internet Accounts preference pane.
- Networks: Could not load Network preference pane.
- Sharing: You can’t open Sharing preferences because it doesn’t work on an Intel-based Mac.
- Users & Groups: You can’t open Users & Groups preferences because it doesn’t work on an Intel-based Mac.
- Parental Controls: Could not load Parental Controls preference pane.
- Date & Time: You can’t open Date & Time preferences because it doesn’t work on an Intel-based Mac.
Any hints on a solution are greatly appreciated!
Best regards, Mikkel

To further explain:
     I didn't need to locate any invisible files in my case. I discovered that I had 4 versions of the System Preferences application on my computer. (The System Preference applicatioin is what loads the preference pains when you seledt System Preferences.) I merely searched for "System Preferences" in spotlight. Spotlight located 4 different versions of the System Preferences application. Some of those versions apparently tied to preference panes that used the old 32 bit encoding from the non Intel processor days. When I clicked on the System Preferences under the Apple menue it was apparently pulling up an old version of the System Preferences applicatioin instead of the correct one in the main Applications folder. I also noted that if I clicked on System Preferernces in the dock or directly in the the Applications folder, I would not have the problem of some of the preference panes not able to be viewed because of the " "it doesn’t work on an Intel-based Mac" message. So I surmised that the correct System Preferences application was in the Applications Folder on the top level of the drive directory like it was supposed to be but that when I clicked on System Preferences from the Apple menue it was loading one of the 3 old System Preferences applications that were buried elsewhere on my hard drive. If you find you have more than one System Preferences application on your hard drive you need to trash the other System Preferences Applications. Before trashing the duplicates, you should test the one in the main Applications folder by directly double clicking it and make sure all the preference panes in it work properly then trash the duplicate System Preferences applications you found in spotlight.
     You may also find like I did, that you not only have duplicates of the System Preferences on your hard drive but also several old duplicates of the entire Applications folder buried on your hard drive. if that is the case you should compare each folder and make sure there is nothing in the old duplicate Applications folders you still need. You will probably find that all of the latest versions of your applications are in the main Applications folder where they should be. If so then trash all the old duplicate Applications folders as well. Keep only the main Applications folder on the top level of the hard drive.
     Trashing all these extra old Applications folders that held old duplicates of all my applications in them on my hard drive solved a bunch of other problems I was having with some of the applications as well.
     For the technically minded: Why does this happen with Mavericks and not with Mountain Lion? I suspect that the old Applications folders and old duplicate System Sreferences applications didn't matter to Mountain Lion because it searched for the System Preference application under the Apple menu differently than Mavericks does. I believe Mountain Lion seached all the top levels of the Application folder before drilling down into any subfolders to find the System Preferences application. Mavericks apparently does it the opposite way. It searches the first subfolder in you Applications folder and drills all the way down in that subfolder first before going to the next subfolder and drilling down in that subfolder all the way to the bottom. So it searches all the subfolders before searching the main level of the Applications folder for System Preferences. Since my old duplicate System Preferences applications where in subfolders with titles that were earlier in the alphabet than the main System Preferences application it searched those subfolders first, found the old System Preferences application in the those earlier folders and loaded the old System Preferences first in the subfolders instead of the correct one further on in the main Applications folder. If any one else has a better idea let me know.
--Scott
micmac2105 wrote:
Hello,
have the same problem, however, I do not understand the hint for solving the problem.
How to show invisible files?
How to find applications/system preferences and call Get info from file menu?
Where are all these things, have never seen them before or heard of them.
Thanks for help
Michael
micmac2105 wrote:
Hello,
have the same problem, however, I do not understand the hint for solving the problem.
How to show invisible files?
How to find applications/system preferences and call Get info from file menu?
Where are all these things, have never seen them before or heard of them.
Thanks for help
Michael
micmac2105 wrote:
Hello,
have the same problem, however, I do not understand the hint for solving the problem.
How to show invisible files?
How to find applications/system preferences and call Get info from file menu?
Where are all these things, have never seen them before or heard of them.
Thanks for help
Michael

How I migrated my caldav data (Contacts and Calendars) to Mavericks + Server

Hello,
I write this post to share with you the steps I had to take for correctly migrating my caldav database (containing data for Addressbook and Calendar Server) from a Lion Server to Mavericks with Server version 3.
The problem occured like this : after downloading and installing mavericks (from the Apple Store), and downloading/installing the Server Utility (version 3), everything migrated just fine, except I had an error message during the first launch of the Server utility, but since it eventually launched, and the services were running, I didn't notice anything at first.
But I rapidly saw a big problem : my caldav data was missing. If a client connected to their calendar application, they would get (after syncing with the server) a new empty calendar named "calendar" (this is typical of a fresh empty caldav database). Same for addressbook clients (empty contacts).
So I searched and read several post that helped me greatly :
https://discussions.apple.com/thread/3890595
https://discussions.apple.com/message/23581552
BUT, what is different in mavericks, is this :
The postgres cluster that hosts the caldav data seems to have move (again!), and this is by exploring several config files that I figured which was the right socket to use to connect to the good postgres cluster.
The clues were in the file : /Applications/Server.app/Contents/ServerRoot/private/etc/caldavd/caldavd-apple.p list
Where I found the precious parameter (under the Postgres section) :
<key>SocketDirectory</key>
<string>/var/run/caldavd/PostgresSocket</string>
What I also figured is that the user account to use for connecting to this postgres cluster is caldav, and not _postgres.
So here are basically the steps I had to take to reinstall my old database.
Step #0 :
First, you need to locate the dump of the old postgres database, I found that the migration process seemed to make a dump inside :
This method ONLY WORKS if you have the following file :
/Library/Server/Migrated/Library/Server/PostgreSQL/Backup/dumpall.psql.gz
If you have that file, you can extract it with the command :
gunzip /Library/Server/Migrated/Library/Server/PostgreSQL/Backup/dumpall.psql.gz
For the rest of the steps I assume this file is named /path/to/your/dump.psql
I think (but I'm not sure of it) that THIS precise dump is different than what you would get inside a manual dump, meaning that in this dump you have all the databases of the cluster. Again I'm not sure, but that's what I think since there are CREATE DATABASE instructions inside it).
If you don't have that file, perhaps you made a dump by hand.
If that's the case, I think that before Step #3 you will have to create an empty caldav database, BUT, I'm not sure the command "calendarserver_bootstrap_database -v" mentionned here will create the caldav in the right cluster (read above). Perhaps we could figured out together if that's your case.
Step #1 : stop the services
sudo serveradmin stop addressbook
sudo serveradmin stop calendar
Step #2 : drop the caldav database
sudo dropdb -h /var/run/caldavd/PostgresSocket -U caldav caldav
If you get this error message :
dropdb: database removal failed: ERROR: database "caldav" is being accessed by other users
DETAIL: There is 1 other session using the database.
then go to Step #2a, otherwise, go straight to Step #3
Step #2a : only if Step 2 didn't work, see above ! (taken from here)
# Connect to the postgres database
psql -h /var/run/caldavd/PostgresSocket -U caldav
# Terminate the current sessions
postgres=> select pg_terminate_backend(procpid) from pg_stat_activity where datname='caldav';
# Exit postgres
postgres=> \q
Step #3 : if your database dump is not from /Library/Server/Migrated/Library/Server/PostgreSQL/Backup/dumpall.psql.gz, then PERHAPS you will have to create manually an empty caldav database (see Step #0)
Step #4 : create the new database from the dump
sudo psql -h /var/run/caldavd/PostgresSocket -U caldav -f </path/to/your/dump.psql> caldav
A lot of things should be displayed, and I eventually got an error at the end :
psql:DataDump.sql:49844: \connect: FATAL: database "collab" does not exist
But in my case that didn't affect my migration.
Step #5 : restart services
sudo serveradmin start calendar
sudo serveradmin start addressbook
I hope that helps you.
Cheers

iCloud adds a new (additional) Address book to Outlook, it contains the iCloud contacts, if the iCloud contacts and the Outlook contacts are the same there will be duplicates:
Look in one Contact Folder only
Delete the Outlook Contacts (you can't delete the whole folder, only the contents
These are your choices.

Mavericks permission issues with Windows Server deduplicated shares.

We have a number of 10.9-10.9.3 - Mavericks - machines installed throughout our facility. Much of the user content is pulled from shares stored on our Windows Server 2012 fileservers with deduplication enabled. I have found that files newly written or unoptimized are able to be accessed without issue - read, written, modified, etc. Once the file gets optomized/deduplicated and Windows adds the P & L attributes - sparse and symlink - the Macs running Mavericks begin to have access issues.
Once the files get deduplicated, users begin receiving read access errors when copying files (see error1 below). This happens when copying to folders within the current folder tree or copying somewhere to the local system. If you 'stop' the copy operation and retry a few more times, it may eventually work for the specific instance but fail again later. I am however, able to copy these files without issue via the terminal.
Other systems running 10.7 do not experience the same issues and are able to access file server resources without issue. Many of the systems having issues are newer and thus not able to be downgraded to 10.8 or 10.7.
I have tried finder replacements such as Pathfinder but the results are the same. I know this is at least similar to the issues many Mac users are already experiencing and posting about but I haven't seen it directly linked to deduplication and the attributes written by Windows server. Has anyone seen this issue? Have any solutions been found?
Error 1:
When copying files after the PL attributes have been set by deduplication.
"One or more items can't be copied to "Foler" because you don't have permissions to read them.
Via the system.log, I am also seeing the following error when accessing these deduplicated file shares. The reparse point tag listed below is "IO_REPARSE_TAG_DEDUP"
Reported error:
"smbfs_nget: filename.ext - unknown reparse point tag 0x80000013"

This is a 'known problem' as discussed on the MacWindows site, see http://www.macwindows.com/More-on-OS-X-conflicts-with-Windows-Server-Deduplicati on.html
It appears to still affect Mavericks 10.9.5 at least from the lack of any positive reports to the contrary. It is perhaps a bit late but it would have been worth signing up for the public beta of Yosemite and testing that and if needed reporting it as an issue for Yosemite as this will be getting more bug fix attention than Mavericks now.
I would agree turning off de-duplication is not a desirable solution, however other possibilities to consider are installing the standard Samba software on a Mac and using that instead of Apple's SMBx software. This is done by using the SMBup installer See http://eduo.info/2013/07/13/smbup-faq it does not officially list Mavericks but I did see at least one report that it still worked. A second possibility is to install ExtremeZ-IP which installs an AFP server on to your Windows Server and also gives the bonus of supporting Spotlight searches.
Note: You should only try SMBup on a test setup first.

Different selection in a single query according to an ID

Hi
I'm looking for a way to perform different selections in a single query according to a specific value:
Here is the first selection:
select g.*,gf.*,gs.*
FROM graphs g
LEFT JOIN graph_frames gf on g.graph_id = gf.graph_id
LEFT JOIN graph_sets gs on gf.frame_id = gs.frame_id
WHERE g.graph_id = :IDHere is the second selection:
SELECT gg.graph_id, gg.graph_name
FROM generic_graphs gg
INNER JOIN generic_graph_frames ggf on gg.graph_id = ggf.graph_id
INNER JOIN generic_graph_sets ggs on ggf.frame_id = ggs.frame_id
WHERE gg.graph_id = :IDNow, the ID cannot be in both the tables and I want to perform that in a single query, UNION cannot be applied since the tables are different.
Any ideas?
Edited by: BluShadow on 14-Sep-2011 09:09
added {noformat}{noformat} tags. Please read {message:id=9360002} and learn to do this yourself.

Example of consolidating the columns...
SQL> ed
Wrote file afiedt.buf
1 with t as (select &id as id from dual)
2 select e.empno, e.ename, e.job, e.mgr, d.deptno, d.dname, d.loc
3 from   (select * from emp cross join t where empno = t.id) e
4         full outer join
5         (select * from dept cross join t where deptno = t.id) d
6*        on (1=1)
SQL> /
Enter value for id: 7521
old   1: with t as (select &id as id from dual)
new   1: with t as (select 7521 as id from dual)
     EMPNO ENAME      JOB              MGR     DEPTNO DNAME          LOC
      7521 WARD       SALESMAN        7698
SQL> /
Enter value for id: 10
old   1: with t as (select &id as id from dual)
new   1: with t as (select 10 as id from dual)
     EMPNO ENAME      JOB              MGR     DEPTNO DNAME          LOC
                                                   10 ACCOUNTING     NEW YORK
SQL>Though, this would be considered poor design because you are trying to query two disperate things, so they should be treated differently. i.e. in my example, I should already know if I'm querying an employee or a department beforehand.

Symlinks treated differently in mavericks stacks

Similar Messages

Maybe you are looking for