Thus Spoke Herman

Right then, finally got round to finishing this.

Need to set up PokerTracker running in Parallels so that it can see the Postgresql DB running on the host machine.

To allow the VM image running in Parallels access to the Mac, I initially went for the host-based networking, thinking that PokerTracker was the only thing and didn’t need an internet connection, but it checks for updates on startup, and I also use Parallels for using Sopcast, which obviously needs internet access, so I had to revert my decision and go back to Shared.

To change the network configuration, just use the Parallels Devices menu and choose the one you want for networking. I disabled the network adapter in Windows and re-enabled to ensure that the new ip address gets set properly.

I opened PokerTracker in Parallels and set up the Postgresql settings in the setting tab - Still didn’t work! I could ping the Mac from the within Parallels, so was a bit confused for a while, but then the penny dropped:

I had to change Postgresql to accept network connections as by default it doesn’t allow non-local connections!

edit postgresql.conf

make sure this line looks like this:

listen_addresses = ‘*’

so it listens on all ip addresses!

edit pg_hba.conf

host all all 10.211.55.0/24 trust

to allow all hosts on this private network to connect! Obviously make this network the one the Parallels DHCP server assigns. This was the default for mine for Shared, it was different for the Host-Based, so check by doing a ipconfig in Windows and an ifconfig on the Mac.

Restart Postgresql once you have changed these files.

I then went into the Postgresql setting in PokerTracker using the new 10.211.55.2 address, which i got from doing a simple ifconfig on my Mac. It works!

I converted a DB, and renamed and set to be the default in PokerTracker. I restarted FullTilt on mac, to see where my hand histories are being saved (~/Documents/HandHistory) and changed this in PokerTracker before taking another Parallels snapshot, so I wouldn’t have to do all this again.

It now works a treat. It was a shame Crossover didn’t work (or I just couldn’t get it working before trying Parallels), but Parallels is pretty cool, and I know now I don’t need any other computer other than this one for my personal use.

NetApp F720 and FC7 (or was it FC9?) from eBay. Another load of hardware I have picked up hugely cheaply as they are now unsupported by their own makers. A habit I just can’t kick…

Now how the hell do they work?

What do we know: the filer head runs a Data ONTAP OS. It is proprietary and I believe needs a license. I have no idea if my filer has one or not.

Got myself a now.netapp account and have access to the documentation.

Using the My Licenses, I search by serial number. Not registered to me. No shit.

I try and register: Registered. Check licenses again: nothing, nothing in my registered products. It did ask for a product number which you get from typing in a command on the filer console. I haven’t even turned it on yet . This is next…

Lets go….

Serial cable by the looks of it. Need to check if it needs anything like a null modem. Need docs… back to now.netapp…
This might be it!

NetApp® Filers
F700 Series
Hardware Guide

actually this talks about LEDs and front diagnostics, lets just power it up and see what it does.

checking sensors
testing memory
finding image
testing all sorts
startup failed
repeat....


i may need to attach the disks and deafen myself further for this.
This entails some seriously loud beeping.
it has found an image though and is loading.
starting - yippee

Initialising PCI devices.


Beeping has stopped. -

Replace NVRAM batteries!
filer 0
Fault on slot 0a


Hmm. Doesn’t look too hot. According to the docs, I need a null modem for the serial console to connect. i only have a 25pin one. wonder if it will work :). Seriously, i have the most ridiculous thing coming out the back on 1 of the servers. Both NetApp and my server are male 9 pin serials. This is the set of connectors:

NetApp M9 - F9 Cable M9 - F9 Adaptor M25 - F25 Null Modem M25 - F25 Adaptor F25 - M25 Adaptor M9 - F9 Adaptor F9 - M9 Server.

Need hyperterminal or some such on my windows server in the rack i am using, but it doesn’t have anything and has no CD Player, hmm, might have to RDC and install it that way.
Add new windows components, click through to Communications and lo and behold, hyperterminal is unchecked. Now there is a security issue with this application on Windows Server 2003 apparently, whcih is why it isn’t installed by default, but I am not that fussed at present. This is no internet facing server, and I might just remove it once done anyway.

Bloody thing won’t do it over network. Had to copy all files to desktop and copy over ones it asked for during install. done now.

Files needed from CD for interest:
HYPERTRS.HL_
HYPERTRS.DL_
HYPERTRS.EX_
HYPERTRS.CH_
HTICONS.DL_
HTRN_JIS.DL_

Serial Settings for a NetApp F720 Filer - 9600 baud, 8 data bits, no parity, 1 stop bit.

We are in. Turning a Power Supply off got me a message. We also have a prompt:

filer>

lets try the command for the registration again:

filer> sysconfig -a
NetApp Release 6.5.6: Tue Oct 4 17:40:11 PDT 2005
System ID: 0016801618 (filer)

Backplane Serial Number: 236634
slot 0: System Board (NetApp System Board V N8)
Model Name: F720
Serial Number: 236634
Firmware release: 2.3_a2
Processors: 1
Memory Size: 256 MB
slot 0: FC Host Adapter 0a (QLogic 2100 rev. 3, 64-bit, L-port, <UP>)
Firmware rev: 1.19.14
Host Loop Id: 7 FC Node Name: 2:000:00a098:039c5a
Cacheline size: 16 FC Packet size: 2048
SRAM parity: Yes External GBIC: No
Link Data Rate: 1 Gbit
0: SEAGATE ST318203FC NA13 17.0GB 512B/sect (LRJ99488000070280Z69)
1: SEAGATE ST318203FC NA13 17.0GB 512B/sect (LRG99899000070481CAM)
2: SEAGATE ST318203FC NA13 17.0GB 512B/sect (LRG781410000704745NF)
3: SEAGATE ST318203FC NA13 17.0GB 512B/sect (LR47258400007052J8K3)
4: SEAGATE ST318203FC NA13 17.0GB 512B/sect (LRG866990000704747B2)
5: SEAGATE ST318203FC NA13 17.0GB 512B/sect (LRB09491000021024LGN)
6: SEAGATE ST318203FC NA13 17.0GB 512B/sect (LRK71932000070295GYY)
Shelf 0: VEM Kernel Version : 2.5 App. Version : 3.1

I/O base 0xbf00, size 0x100
memory mapped I/O base 0x84000000, size 0x1000
slot 0: SCSI Host Adapter 0b (Qlogic ISP 1040B, <ONLINE>)
Fast/Wide, Differential
Firmware Version 4.66.00 Clock Rate 60MHz.
I/O base 0xaf00, size 0x100
memory mapped I/O base 0x84200000, size 0x1000
slot 0: Ethernet Controller e0
MAC Address: 00:a0:98:00:82:68 (auto-100tx-fd-up)
I/O base 0xae80, size 0x80
memory mapped I/O base 0x84201000, size 0x400
slot 9: NVRAM (NetApp NVRAM II)
Revision: B3
Serial Number: 24402
Memory Size: 8 MB
Battery 1 Status: 0% (0.01v)
Battery 2 Status: 100% (3.08v)
memory mapped I/O base 0x84001000, size 0x80
I/O base 0xbe80, size 0x80
memory mapped I/O base 0x80000000, size 0x4000000
filer>


much stuff. it also seems to have recognised all the 7 17G seagate disks in my FC7 array.

Battery 1 status 0%
Battery 2 status 100%

This explains the front panel messages, but why is it still knackered?

Not attached to network yet, lets see what happens when i do this.
cool, it has a bash like command history!
right, ethernet is now up, but no ip.

filer>ifconfig e0 192.168.0.13

works. now pingable.

filer>license
cifs CNDWHEM

There are a lot of options/extras and other cool looking things, but they are all unlicensed. At least cifs has a working license, otherwise this is a paperweight, as there is no way I am paying for one of those license things (if I could even get one).

Lets see what we have under cifs then:


filer> cifs shares
Name Mount Point Description
---- ----------- -----------
ETC$ /etc Remote Administration
BUILTIN\Administrators / Full Control
HOME /vol/vol0/home Default Share
everyone / Full Control
C$ / Remote Administration
BUILTIN\Administrators / Full Control
filer>


added myself as a user

filer> useradmin useradd matt
New password:
Password must contain at least 2 alphabetic characters and
at least 1 numeric or special character
New password:
Retype new password:
User <matt> added.
filer> Thu May 1 21:51:51 GMT [rc:info]: User <matt> added


Oh look, the date is wrong. I will fix if I can get everything else sorted.
Connected as that user from a different machine, whoa, we are in! We now have a working filer! It has a df command and it only shows 11G though:

filer> df -h
Filesystem total used avail capacity Mounted on
/vol/vol0/ 11GB 155MB 11GB 1% /vol/vol0/
/vol/vol0/.snapshot 3056MB 14MB 3042MB 0% /vol/vol0/.snapshot


A problem message popped up in the console, which was the “Fault on slot 0a” issue, this was rectified when I attached another power supply to the disk array. Greedy bastard wants 2. The noise suddenly reduced massively and the it even reported that the fault had now been rectified. Ho Ho Ho. Just need to work out how to replace NVRAM batteries now, and work out where all the disk space is


filer> shelfchk
Only shelves attached to channel 0a should have all LEDs ON.
Are these LEDs all ON now? yes
filer> Thu May 1 22:07:14 GMT [rc:info]: Disk Configuration - No Errors Identified

Disk array is fine. It seems I need to create a new volume. Back to docs for a bit…
Right, I need to see what disks are available for a new volume:

filer> vol status -s

Spare disks

RAID Disk Device HA SHELF BAY CHAN Used (MB/blks) Phys (MB/blks)
--------- ------ --------------- ---- -------------- --------------
Spare disks for zoned checksum volumes only
spare 0a.2 0a 0 2 FC:A 17000/34816000 17366/35566480
spare 0a.3 0a 0 3 FC:A 17000/34816000 17366/35566480
spare 0a.4 0a 0 4 FC:A 17000/34816000 17366/35566480
spare 0a.5 0a 0 5 FC:A 17000/34816000 17366/35566480
spare 0a.6 0a 0 6 FC:A 17000/34816000 17366/35566480
filer>
Cool. Lets get creationist on it!

I went for 4 as I got a RAID warning about not having any spares so thought I might get add the 5th later

filer> vol create vol_main 4
Thu May 1 22:38:56 GMT [raid.vol.disk.add.done:notice]: Addition of Disk /vol_main/plex0/rg0/0a.5 Shelf 0 Bay 5 [SEAGATE ST318203FC NA13] S/N [LRB09491000021024LGN] to volume vol_main has completed successfully
Thu May 1 22:38:56 GMT [raid.vol.disk.add.done:notice]: Addition of Disk /vol_main/plex0/rg0/0a.4 Shelf 0 Bay 4 [SEAGATE ST318203FC NA13] S/N [LRG866990000704747B2] to volume vol_main has completed successfully
Thu May 1 22:38:56 GMT [raid.vol.disk.add.done:notice]: Addition of Disk /vol_main/plex0/rg0/0a.3 Shelf 0 Bay 3 [SEAGATE ST318203FC NA13] S/N [LR47258400007052J8K3] to volume vol_main has completed successfully
Thu May 1 22:38:56 GMT [raid.vol.disk.add.done:notice]: Addition of Disk /vol_main/plex0/rg0/0a.2 Shelf 0 Bay 2 [SEAGATE ST318203FC NA13] S/N [LRG781410000704745NF] to volume vol_main has completed successfully
Thu May 1 22:38:57 GMT [config_async_0:notice]: Volume vol_main has been added to the system.
Creation of a volume with 4 disks has completed.
filer> Thu May 1 22:38:57 GMT [download.update:info]: Begin bootblock update, prototype is 0a.1


But when I checked, I saw that 1 had been set as a parity one anyway:

filer> vol status vol_main -r
Volume vol_main (online, raid4) (zoned checksums)
Plex /vol_main/plex0 (online, normal, active)
RAID group /vol_main/plex0/rg0 (normal)

RAID Disk Device HA SHELF BAY CHAN Used (MB/blks) Phys (MB/blks)
--------- ------ --------------- ---- -------------- --------------
parity 0a.2 0a 0 2 FC:A 17000/34816000 17366/35566480
data 0a.3 0a 0 3 FC:A 17000/34816000 17366/35566480
data 0a.4 0a 0 4 FC:A 17000/34816000 17366/35566480
data 0a.5 0a 0 5 FC:A 17000/34816000 17366/35566480


Rather than have a wasted disk (do I really need a spare, not at present. Maybe I will undo later and have a spare once I work out what to do with this insanely noisy thing) - I need to destroy and start again:


filer> vol offline vol_main
Thu May 1 22:41:41 GMT [config_async_0:notice]: CIFS - disabled for volume vol_main
Volume ‘vol_main’ is now offline.
filer> vol destroy vol_main
Are you sure you want to destroy this volume? yes
Thu May 1 22:41:51 GMT [raid.config.vol.destroyed:info]: Volume ‘vol_main’ destroyed.
Volume ‘vol_main’ destroyed.
filer>

filer> vol create vol_main 5

WARNING! Continuing with vol create will result in having no spare disk available for one or more raid groups.
Are you sure you want to continue with vol create? yes
Thu May 1 22:44:03 GMT [raid.rg.spares.low:warning]: /vol0/plex0/rg0
volume has been created with 1 disk added to the volume. 4 more disks need
to be zeroed before addition to the volume. The process has been initiated
and you will be notified via the system log as the remaining disks are added.
filer>
filer> Thu May 1 22:45:00 GMT [monitor.globalStatus.nonCritical:warning]: There are not enough spare disks.


Bugger


filer> vol status
Volume State Status Options
vol_main restricted raid4
initializing
vol0 online raid4 root, maxdirsize=5242
filer> vol offline vol_main
vol offline: Invalid operation on volume vol_main being created.
filer> vol destroy vol_main
vol destroy: Invalid operation on volume vol_main being created.
filer>

Have I shafted it already?

filer> vol status vol_main -r
Volume vol_main (restricted, raid4, initializing) (zoned checksums)
Plex /vol_main/plex0 (offline, empty, active)

Targeted to volume but not yet assigned to a raid group
RAID Disk Device HA SHELF BAY CHAN Used (MB/blks) Phys (MB/blks)
--------- ------ --------------- ---- -------------- --------------
pending 0a.2 0a 0 2 FC:A 17000/34816000 17366/35566480 (zeroing, 21% done)
pending 0a.3 0a 0 3 FC:A 17000/34816000 17366/35566480 (zeroing, 22% done)
pending 0a.4 0a 0 4 FC:A 17000/34816000 17366/35566480 (zeroing, 20% done)
pending 0a.5 0a 0 5 FC:A 17000/34816000 17366/35566480 (zeroing, 22% done)
pending 0a.6 0a 0 6 FC:A 17000/34816000 17366/35566480


Phew, not quite. The warning must be that I have just used all the disks.

This is taking quite some time. Once done I reckon I need to create a new cifs share and away we go! I can then do some tests on the speed of this end-of-life ebay gloriousness!

filer> vol status vol_main -r
Volume vol_main (online, raid4) (zoned checksums)
Plex /vol_main/plex0 (online, normal, active)
RAID group /vol_main/plex0/rg0 (normal)

RAID Disk Device HA SHELF BAY CHAN Used (MB/blks) Phys (MB/blks)
--------- ------ --------------- ---- -------------- --------------
parity 0a.3 0a 0 3 FC:A 17000/34816000 17366/35566480
data 0a.6 0a 0 6 FC:A 17000/34816000 17366/35566480
data 0a.2 0a 0 2 FC:A 17000/34816000 17366/35566480
data 0a.5 0a 0 5 FC:A 17000/34816000 17366/35566480
data 0a.4 0a 0 4 FC:A 17000/34816000 17366/35566480

filer> df -h
Filesystem total used avail capacity Mounted on
/vol/vol0/ 11GB 155MB 11GB 1% /vol/vol0/
/vol/vol0/.snapshot 3056MB 14MB 3042MB 0% /vol/vol0/.snapshot
/vol/vol_main/ 47GB 1508KB 47GB 0% /vol/vol_main/
/vol/vol_main/.snapshot 11GB 0GB 11GB 0% /vol/vol_main/.snapshot
filer>


Created fine, lets get it shared via cifs:

filer> cifs shares -add data /vol/vol_main
filer> cifs shares
Name Mount Point Description
---- ----------- -----------
ETC$ /etc Remote Administration
BUILTIN\Administrators / Full Control
HOME /vol/vol0/home Default Share
everyone / Full Control
C$ / Remote Administration
BUILTIN\Administrators / Full Control
data /vol/vol_main
everyone / Full Control
filer>


And a quick test from another machine works fine. We are now ready to see how fast this goes, and this will have to be another time, as I need to get some sleep as my day tomorrow has admin all over it, and it is late already…

filer> halt -t 0.1
CIFS local server is shutting down...
CIFS local server has shut down...
Thu May 1 23:56:22 GMT [kern.shutdown:notice]: System shut down because : “halt”.
ok


I finally got a chance to do some work at the server without the pain of using Remote Desktop. I was messing around a bit, and needed a restart. Restart I did. However, after the initial Windows Server 2003 screen, I got the infamous blue screen of death:

“The registry cannot load the hive (file)”
“SystemRoot\System32\Config\SOFTWARE”

Now thats a failure! Hive file, WTF? I didn’t get a screen shot, but some other dude has had the same thing. I actually didn’t wait for the memory to dump to disk, I have no time for niceties.

Luckily my method of sorting this mess out was less tortured than his:

http://content.zdnet.com/2346-10741_22-39528.html

I tried Safe Mode. No cigar, same screen. Bugger.

I have the Windows Server 2003 CDs so I got them out. 1 issue. No CD Player in the server and I do not have a USB CD player, just IDE ones. Bugger.

Get the server out of the rack, take the hood off and connect the CD drive to the spare IDE slot and connect (just reached) to the spare molex. Was not recognised by motherboard. Bugger.

Swapped jumper on CD Drive to cable select - was on master. We are off and running. Changed the BIOS to try the CD drive first. We are into the Windows installer! When given the option, I press R for the Recovery Console. After being prompted for the admin password, I ran the following commands:

cd system32\config
ren software software.bak
copy c:\windows\repair\software

This goes to the directory where the dodgy registry software crap is, takes a copy (just in case) and then copies a repair, older version from a repair directory. Reboot time…

exit

It works!
Shutdown, remove IDE CD player, close hood. startup. It works!
Put back in rack
It works!
Wow. I can now do some work!

I got a copy of Office 2003, and installed using CrossOver again. I used the existing Microsoft Office 2003 installation rather than the custom one this time. After downloading many fonts, I just chose to install Access. This did not take very long. I could not create and save a new blank database though. PokerTracker would also still not start. Without access even running, I might have to use a VM image.

I decided to cut my losses and install Parallels and give it a go this way. I don’t want to dual boot, so Parallels seems a good option. I downloaded and installed and have 15 days to evaluate. I used Parallels Transporter to install a Windows VM from a Windows 2003 server installation I have running on a server tucked away in a dusty corner of the basement. It took a very long time. Once installed, I installed PokerTracker and took a snapshot. I also got myself a PokerTracker license, so I could use the postgres stuff already installed on the Mac. I activated PokerTracker and took another snapshot (also installed SopCast as a bonus before taking the snapshot, which also didn’t work fully in CrossOver). The issues now remaining are how to access the postgres DB running on my host Mac and how to access the directories of the FullTilt application, also running on the Mac. I want the DB on the host so I can simply use a Parallels snapshot and just restart the snapshot each time for obvious reasons. I also want the DB available outside of Parallels, as the less I have running in Windows the better.

I haven’t gotten PokerTracker talking to the Postgres DB on the host yet though. I am posting as I haven’t had time to look at this in a while and believe the longer things stay in draft mode, the less likely anything will ever happen….

conclusion of this saga can be found here

Download the pokertracker installer and install using the unsupported software installation for crossover. I want to use the postgres database server as the database as I have always hated Access and getting that to work with crossover has had mixed reviews according to some forum posts I have seen, although crossover lists various Access versions as supported.

To install postgres I used the instructions on the developer.apple site - http://developer.apple.com/internet/opensource/postgres.html

I need to install fink (http://www.finkproject.org/download/index.php) as well. I am using Leopard, and at the time of writing, there is no binary, so I need to get the source. After downloading I realised I had no c compiler installed! A handy error message told me I needed the developer tools:

....
Checking cc... not found.
ERROR: There is no C compiler on your system. Make sure that the Developer
Tools are installed.

Got the XCode3 stuff from http://developer.apple.com/tools/download/ and just installed the XCodeTools.mpkg package. This took just under 10 minutes to install, most of the time seemed to be installing documentation, which is nice. Seems a bit strange that a Mac doesn’t come with any compiler, but I am used to Linux when it comes to non-Windows machines, and this is my first Mac.

I have used WebObjects before and had no interest in that, so left it out. Reminds me too much of a sketchy project I worked on in another lifetime.

Once the dev tools are all installed, I can run the fink ./bootstrap command again. Unsurprisingly, I went for the defaults for most of the options, although I did turn on the unstable option as I like new broken stuff and I went for UK based mirrors…
I then was told to make a cup of tea in the INSTALL notes, so I did. At 23.58. Lets see how long this takes
… good job i waited as first curl call timed out happily downloading from another mirror.

I also notice, as I sit back down that my MacBook Pro is making more fan noise than ever. Hardest its worked so far methinks. At least it now knows what is ahead of it.

While this was grinding away, I downloaded the PostgreSQL source, as there is obviously no binary for OSX. I went for the 8.3 version and seeing as I saw bzip2 flash past on this fink thing, I went for the postgresql-8.3.0.tar.bz2 source files.

Fink finished at 00.11. Might be a late one, but I am warming to this task I think.

Now the instructions say install the readline using fink, but this happens:

me:fink-0.28.0 me$ sudo  /sw/bin/fink install readline
Password:
Scanning package description files..........
Information about 268 packages read in 0 seconds.
Failed: no package found for specification 'readline'!

This sucks. I added /sw/bin to my $PATH to make things easier and had a look at the other options for fink. A list was what i was after and this gave me:

me:fink-0.28.0 me$ fink list readline
Scanning package description files..........
Information about 268 packages read in 0 seconds.
 p   term-readline-pm588                      [virtual package]

This doesn’t look right. As readline is a GNU app, I wondered if the gnu mirror was cocked up, so went looking for a conf file rather than having to rerun the configuration again. It is /sw/etc/fink.conf and all the mirrors are listed there. As the terminal window does not have infinite window buffer size, the other mirror options had vanished into terminal nirvana, so I checked the gnu site for different mirrors (I could just download the readline source from here, but I am determined now). I added Mirror-gnu: ftp://ftp.cwru.edu to the list. This did sweet FA. I ran a fink selfupdate and watched the Dual Core go. I did have a lot of mirror trouble it seemed. Not sure if my connection is crapping out or if they are just unavailable. Got one in the end.

me:fink-0.28.0 me$ fink list readline
Information about 6504 packages read in 1 seconds.
     readline               4.3-1028          Comfortable terminal input library
     readline-java          0.8.0-1024        Readline support for Java
     readline-shlibs        4.3-1028          Comfortable terminal input library
     readline5              5.0-1005          Comfortable terminal input library
     readline5-shlibs       5.0-1005          Comfortable terminal input library
     term-readline-gnu-pm5  1.15-1004         Perl extension for the GNU Readline
     term-readline-pm       1.0203-12         Minimal interface to Readline
 p   term-readline-pm588                      [virtual package]

Cool. Went for the 5 version. More Core Duo action.

On to Postgres, files downloaded, we are nearly ready to crack on with this latest nonsense:

me:Downloads me$ md5 postgresql-8.3.0.tar.bz2
MD5 (postgresql-8.3.0.tar.bz2) = 53d6816eac7442f9bc8103439ebee22e
me:Downloads me$ cat postgresql-8.3.0.tar.bz2.md5
MD5 (postgresql-8.3.0.tar.bz2) = 53d6816eac7442f9bc8103439ebee22e

Rock n Roll. I can feel the tiredness of getting up early because of a timezone mistake and an evening boxing training taking its toll.
The INSTALL document has a short install instructions at the top. I will go through it and only put stuff here that deviated from the expected. I have a feeling old Core Duo is going to get another pounding.

First up, we need to do some extra includes for the ./configure, as mentioned in the apple.developer doc /configure –with-includes=/sw/include/ –with-libraries=/sw/lib
Second, I don’t have gmake, I have make so I used that instead.
Third, I forgot to sudo the make install, so got permission problems. If that is the only cock-up left in me tonight I will be very happy. Install completed pretty quick..
Some graphical UI stuff next according to apple.developer doc. I added the postgres user. When this was done, I could create a data directory for the db and then initialise it:

me:postgresql-8.3.0 me$ sudo sh
sh-3.2# mkdir /usr/local/pgsql/data
sh-3.2# chown postgres /usr/local/pgsql/data/
sh-3.2# su -l postgres
me:~ postgres$ /usr/local/pgsql/bin/initdb -D /usr/local/pgsql/data/
The files belonging to this database system will be owned by user "postgres".
This user must also own the server process.

The database cluster will be initialized with locale C.
The default database encoding has accordingly been set to SQL_ASCII.
The default text search configuration will be set to "english".

fixing permissions on existing directory /usr/local/pgsql/data ... ok
creating subdirectories ... ok
selecting default max_connections ... 20
selecting default shared_buffers/max_fsm_pages ... 2400kB/20000
creating configuration files ... ok
creating template1 database in /usr/local/pgsql/data/base/1 ... ok
initializing pg_authid ... ok
initializing dependencies ... ok
creating system views ... ok
loading system objects' descriptions ... ok
creating conversions ... ok
creating dictionaries ... ok
setting privileges on built-in objects ... ok
creating information schema ... ok
vacuuming database template1 ... ok
copying template1 to template0 ... ok
copying template1 to postgres ... ok

WARNING: enabling "trust" authentication for local connections
You can change this by editing pg_hba.conf or using the -A option the
next time you run initdb.

Success. You can now start the database server using:

    /usr/local/pgsql/bin/postgres -D /usr/local/pgsql/data
or
    /usr/local/pgsql/bin/pg_ctl -D /usr/local/pgsql/data -l logfile start

Oh man we are getting close now!

me:postgresql-8.3.0 me$ sudo su - postgres
me:~ postgres$ echo $PATH
/bin:/usr/bin:/usr/sbin:/sbin:/usr/local/bin:/usr/X11/bin:/usr/local/pgsql/bin:/sw/bin
me:~ postgres$ pg_ctl -D /usr/local/pgsql/data -l logfile start
server starting
me:~ postgres$ createdb pokertracker
me:~ postgres$ psql pokertracker
Welcome to psql 8.3.0, the PostgreSQL interactive terminal.

Type:  \copyright for distribution terms
       \h for help with SQL commands
       \? for help with psql commands
       \g or terminate with semicolon to execute query
       \q to quit

pokertracker-# \q
me:~ postgres$

Started up the PokerTracker application through crossover and lo and behold, it fails with:

unable to connect to the PTREG2 database

Goddamnit, is it trying to connect to Access DB first. If I remember correctly I think it does, and you have to convert the DB to be a Postgres one. Shite. It is now 1.30am and I need sleep. Next post is how I got hold of a version of Access 2000 :).