Carlson Skunk Works

How about that? Less than a month since my last post!

Progress is being made and I now have something that most people would recognize as an airplane part. The fin is basically finished! There is still some trim work and sanding to be done, and then I will need to seal it, but the assembly is complete.

Last post I mentioned that I had been in San Francisco for Oracle OpenWorld. I found a couple of pictures of the area where the conference was held, so I thought I would post them. These were taken from the overpass above Howard Street. The first one is looking to the west and shows the corner of Moscone West. The second picture is facing east from the same location and shows the tent that covers Howard Street and some of Moscone South.

Now back to the airplane.

These two pictures show how the fin skin and gusset strip will fit on this side of the fin. I didn’t have the 0.050 steel sheet called for in the plans to make the bracket for the elevator wires to fasten to, so I made a tag board version of the bracket and used that to locate and size a hole in each of the fin gusset strips. The holes will allow me to install the bracket once I make it. In the mean time, I can complete the fin and seal it properly.

This picture shows the hole in more detail. The hole is 42mm in diameter and is centered on the center line of the bolt that fastens the bracket to the fin. The rear edge of the hole is at the leading edge of the fin spar.

I was too tired to tackle gluing up the fin plywood, so I moved on to the hinge plates. On Thursday I stopped at the Marlatt Machine Shop to see if they would happen to have some of the metal for making the various fittings. Gary did not have the 4130 sheet steel that I needed, but he did have the aluminum for the fin fittings.

The first thing to do is to make a template of the hinge plate. The hinges for the elevator and the rudder are identical, so I need to make 14 hinge plates, two for each hinge. When I made the template, I was a bit confused by the markings for the top edge of the hinge plate. Looking at the drawing one way, it appeared that the top edge was to be 12mm long, 6mm each side of the center line. When I cut that template out and placed it on the plans that top edge just didn’t line up correctly, so I re-did the template so it now has an 18mm top edge, 9mm each side of center.

In order to waste as little material as possible, I made a second template that extends the angled sides to their max. This resulted in an equilateral triangle.

Here the hinge plate template sits on top of the triangle.

Enough for one night. I was tired and decided to head to bed.

Sunday afternoon found me much better rested and well fed after a great lunch of pork roast, mashed potatoes and squash. I think we are practicing for Thanksgiving already. After lunch everyone else decided to take a nap, so I changed clothes and headed to the basement to glue the plywood to the fin.

Here are two pictures showing the finished fin. I added the clamps to the leading edge to ensure that the plywood would fit around the edge as much as possible. I had sanded the leading edge enough that the plywood easily contacted lots of the leading edge strip, but I just wanted to be sure that it fit as well as possible. I will leave the clamps in place for a couple of days to ensure that the glue is fully cured before I take them off.

Once the fin was fully glued, I moved to the garage to start cutting hinge plates.

The hinge plate material is 2024-T3 aluminum plate, 0.063″ thick. I had Gary cut two pieces of aluminum for me that were 2.5″ X 36″. These will make the hinge plates plus several other pieces. I started by laying the hinge plate template in place and tracing around it with an ultra-fine point permanent marker. Then I place the triangle template over that and marked the long edge so that I could have a full length line to follow.

My wife insisted that I purchase a new hacksaw for this project. She is so great about the airplane! She even paid for the saw! Anyway, the new saw made cutting the aluminum very easy.

Once I had the first hinge plate cut, I would start each hinge plate by using the triangle template.

Then I positioned the hinge plate template inside the triangle and marked the other three sides.

Then I used a piece of 2X4 and a clamp to hold the metal in position while I cut it with my new hacksaw.

First I cut the 14 triangles, then I went back and cut the corners off of each of the triangles. Once I had the hinge plates rough cut I used my file to smooth the edges a bit. That way the plates sit together very nicely and I reduce the chances of cutting myself on a rough edge.

I took some packing tape and wrapped the hinge plates so I could stack drill them. I temporarily taped the hinge plate template to the top of the stack and marked the locations of the holes with a center punch. Then I held the stack in my drill press vice and proceeded to drill the three holes.

The holes were drilled using a 3/16″ drill. When the first hole had been drilled, I put one of the AN-3 bolts through it to ensure that the hinge plates would stay positioned correctly. Then I drilled the second hole and put another bolt through it. Finally, I drilled the third hole. Of course, I forgot to take pictures of this process and the finished parts. Oh well, I will try to remember to have those in my next post.

The plan is to clean up each of the hinge plates as needed, then bolt the whole stack together and file the edges so that they are all the same. I cut each of them a bit oversize to allow for this. I will include pictures of this process in my next post.

It has gotten to be a long time since my last post! Fortunately, there has been a fair amount of progress made on the airplane as well.

At this point the spars are basically done. I haven’t gotten any metal for the various fittings, so I have been doing as much as I can without the fittings. Therefore, I still need to finish the root end of the front spars and they all need a final sanding.

So to get everyone up to date, here are the rest of the pictures that I have taken of building the spars. Here is a picture of how I clamped the plywood after it was glued up. I figured that it would save time to glue up both of the rear spar webs at the same time.

Now it is on to the tail feathers!

This picture shows the fin spar in the foreground and the rudder spar toward the back. The fin ribs are laying across the two spars, ready to be glued to the fin spar. The top three ribs needed to be glued at right angles to the spar.

Here is a picture of some of the parts for the rudder ribs. They are not glued up in this picture, but they have now been glued in the first step toward making the rudder ribs.

I got so excited about actually assembling the fin that I forgot to take any pictures until I got this far.

At this point my wife said that the fact that we are really building an airplane sunk in. The parts are now starting to look like airplane parts to her.

The length of the fin spar is not quite 60 inches. This is about the position that the fin will be in when the plane is actually flying. Just imagine it with it’s fabric covering, the rudder trailing it and the rest of the plane ahead of it.

Last week I was in San Francisco at Oracle OpenWorld. It was a good time and I learned a lot, but nothing got done on the plane while I was gone.

Today I was able to get back to my lunch routine of doing at least a little bit on it. I don’t have any pictures yet, but I scared and glued up the plywood for the skins and gussets for the fin. I am hoping to be done with the fin next week and ready to move on to the rudder.

Well, the past couple of weeks has seen the assembly of the wing rear spar frames and the webs for the rear spars. Therefore, progress continues to be made.

The webs for the rear spars were scarfed last week and the assembled spar frames were sanded. This was a bit of a project since my wife has banned any further sanding in my basement work room. Therefore, I had to bring the spars up to the garage for sanding, then clean them off and bring them back to the basement.

Here are some pictures of the wing rear spars during construction:

This is the long view showing everything ready to be glued. I built two of the spar frames like this and will apply the webs to opposite sides so that I end up with right and left spars.

This shows the filler block for the jury strut attachment.

This is the root end of the assembly. I put a stop block at this end to provide a good reference point for other measurements.

Here is the center of the spar where the strut will attach. I cut the filler block a bit long and ended up having to fit the compression strut fillers at rib position 8.

Of course, the tip end of the spar. I ended up using a piece of left over 9mmX18mm material from the front struts to make the end of the spar. Once the glue had cured I sanded it down so that it is now level with the rest of the spar.

Once both of the rear spar frames had been glued and the glue allowed to cure, I cleaned off all of the wood blocks from the spar table. I then took the web material and laid it out for gluing. I left the parchment paper and one layer of wax paper on the table. I ran one web down the left side of the table and the other down the right side of the table.

Before I scarfed the web pieces I marked them so that I would know which pieces went together and in what orientation. That way I made sure the grain was running in the right direction.

I went over to Menard’s and picked up a bunch of clamps. Then I started at the root end and clamped the first piece of plywood to the table, scarf side up to accommodate applying glue. I worked on both webs at the same time, so the joints lined up as I went along.

Then I fitted the next piece of plywood so that the scarf joint was nice and level and smooth. At this point I clamped the second piece of plywood to the table. Next I lifted the end of the second piece of plywood and applied the glue to the joint. When I had glued the first joint of both webs, I laid a piece of wax paper across them, put a short board across the table and clamped the board over the joints. This should apply an even pressure across both scarf joints.

Then I moved on the next joint and continued working my way toward the tip end of the webs. As I moved down the web I moved any extra clamps along with me. I tended to use lots of small clamps to keep the web pieces from moving while I glued the joints. I ended up with four scarf joints in each web.

Unfortunately, I forgot to take any pictures of the process. I will try to get a picture of the fully clamped webs for the next post.

Progress is continuing to be made. Steady by jerks as my grandfather used to say. However, doing a little bit each day during my lunch break is really helping to keep things moving.

The two front spars for the wings are now complete and the two rear spars are in process.

Sunday afternoon I spent a few hours cutting and planing the Douglas fir parts for the rear spars. All the material is now ready at hand.

However, the spar table still needed to be cleared off from building the front spars and prepared for building the rear spars. Therefore, my Monday lunch time was spent clearing the table. I pulled the blocks that held parts in place, removed the wax paper and the parchment paper. New parchment paper was rolled out and stapled to the table and a reference line was drawn the full 16 feet of the table. Then on Tuesday I drew out the rest of the spar full size on the parchment paper and overlaid it with wax paper.

The table now has most of the part for the first spar on it and I have stapled blocks of wood to keep things in place until I can glue them up. Hopefully I will be able to get some pictures of the spar as I build it, but I haven’t had a chance to do so yet.

Here are some pictures of the front spars as the plywood web was installed. As noted in my last post, there are four pieces of 2mm plywood for each web. The middle joint was lined up with the filler blocks where the struts will attach to the wings, therefore, I used a but joint in this location as there is a lot of solid wood on each side of the joint. The other two joints on each wing were scarf joints. The next few pictures show how I laid out the but joint and the finished joint.

First I marked the center of the filler block:

Then I laid one half of the web in place so that it crossed the center mark and marked the end of the web:

Here you can see how much wood is on each side of the joint:

Then the glue is applied, the web is positioned and I stapled everything LOTS!

Here is the scarfed joint toward the root end of the rib, notice that most of this joint just happens to fall across the filler block for the jury strut.

Here is the scarfed joint toward the wing tip. This joint has less support from filler blocks.

All the stapes were removed and this surface scuff sanded before gluing up the other half spar and clamping it in place. The scarf joints required more sanding than the rest of the web just to get all the excess glue off.

I will try to get some detail pictures as I build the rear spars and put them in the next post.

Here is a special picture to round out this post. Saturday was the Antique Aircraft Association local fly-in at the Minden, NE airport. Our EAA chapter was asked to serve breakfast and lunch, so I spent the day watching airplanes. One of the planes is this PT-19 owned by my flight instructor, Ed Nelson of Nelson Aviation in Sidney, NE.

Well, I made a bit more progress this morning. Yup, this MORNING! My son had to be at the hospital at 4:00 AM to do a job shadow thing and, since my wife works close to the hospital, she went with. By the time they left I was awake and the coffee was ready.

What to do? Grab a cup of coffee and build airplanes!!!

All the half spars are basically done, but there are a couple that still need a few parts to be glued on. That is where I started. I pulled the finished half spar from the jig and replaced it with another one that still needed the spacer blocks for the compression struts and the aleron pully mount. I fitted those parts, then turned my attention to making my first ever scarf joints.

I had already cut the 2mm plywood strips for the spar webs, but they needed to be scarf joined to turn them into a single piece of plywood for the spar web. Scarfing is simply tapering the ends of the plywood at about a 12:1 taper. Since the plywood is 2mm thick, my scarf joints will be aobut 24mm wide.

On each piece I measured back from the edge of the plywood 25mm and drew a line to show where I was going to do the scarfing. Then I grouped the pieces by which end and angle the scarf would have to be made. I had two groups of four pieces of plywood each. The pieces were stacked such that they were back 25mm from the edge of the piece immediately below. Once in position I clamped them together and proceded to make the taper.

I started with a block plane and roughed them down at least half way. Then I switched to a sanding block and 60 grit sandpaper and proceded to make dust until the taper looked pretty good (to me anyway).

Each sheet of 2mm plywood is comprised of three layers of wood. The middle layer has its grain running perpendicular to the two outside layers. As the sanding removes the wood the layters show up as different colors. With each joint being 24mm wide, each layer of wood will show up as a color band about 8mm wide. The surface of the scarf ideally should have all parallel lines. I think I came reasonably close.

Here are a couple of pictures of my efforts:

The only problem that I had was that one of the pieces got inserted into its group backwards and I ended up scarfing the wrong end of that one piece. This is not a big problem because I can just scarf the other end of this piece and re-cut the end that I scarfed by mistake.

Once I shook all the dust off of the pieces and myself, I selected a couple of pieces to be my first ever attempt at gluing up a scarf joint.

The final step for the morning was to put wax paper underneith all the joints that I am going to glue. Mix up a batch of T88 glue and start smearing it on. The parts on the half spar are pretty easy to do ( I have now done quite a few of these). Then I lay a piece of wax paper over each joint and drive in a staple or two to hold things in place until the glue cures. I also laid a piece of wax paper under the scarf joint and proceded to smear the joint with glue.

Getting the plywood to lay still while I adjust it to be straight was a bit tricky. I finally put another piece of wax paper over the joint and put a staple through the joint near one edge. Then I took my long straight edge and made sure the two pieces of plywood were straight with each other and put a staple through the joint near the other edge. Then I added a couple more staples in the middle to make sure that everything was being held securely and evenly. I also did a second scarf joint.

Later today (lunch time) I will put the last piece of fir on the last of the half spars and glue up the other two scarf joints.

Well it has been quite a while, but progress has been made. Jeremiah and I found some time to rip the strips required for the wing spars and plane them to size. Then I finished the spar table, covered it with parchment paper and drew out a front spar full size. Then I laid on some of the strips that we had cut and fastened wood positioning blocks to hold things in place.

I now have four of what I am calling “half spars”. The next step is to cut the 2mm plywood web strips.

I also ordered a formal Airplane Builder’s Log book which should be arriving any day now. It is coming USPS, so it could be a few more days.

The fitting of the web pieces is a bit confusing, so I called Jerry for some advice. He thought that my plan for fitting the web would work very well. I will add more detail of that step when I get to it. For now I have some pictures of the last

Well, here is my first article on the topic of woodturning. It has been well over ten years since I have attempted to turn irregular shaped wood into round wood. Therefore, I would classify myself as a beginner.

So far I have checked out some books from the library and read them completely. I have located several blogs regarding this subject and have gone through them pretty completely. I purchased the book by Keith Rowley and am reading through that VERY carefully. I am part of the TurningAndTalking email group.

The equipment that I have is fairly limited, but should get me started.

I obtained a lathe via barter, so there was no expense there. However, I did need to buy some material and build a stand for it. Here are some pictures of the lathe and stand:

Here is the finished stand with the lathe mounted on it. The lathe is held onto the stand with six four inch lag bolts.

This is a pretty basic lathe. It does not have a Morse taper on either the headstock or the tailstock. It does, however, have a live center on the tailstock and a 1×8 thread on the headstock. I slid the tailstock up to the headstock and the points met! That was a real surprise! I had expected that I would need to do a bunch of shimming or grinding or fitting to be able to use it.

Here you can see that I have unscrewed the spur drive from the headstock. There is a faceplate that screws onto the headstock in place of the spur drive. With the 1×8 thread on the headstock I might be able to justify getting a four jaw chuck at some point in the future. Well, I can dream can’t I???

This unit has four speeds that are chaged via moving the belt across the pully sets. This could be the most difficult part of this project. There is no easy way to loosen the belt for movement. The screw holding the motor in place needs to be loosened by means of a socket and extension. The lowest speed is listed as 1100 rpm and the current setting is 1600 rpm. If I start by doing some work between centers I should be ok where it is for a while. I am sure it won’t be too long before I will need to learn to adjust the belt.

I also purchased a used bench grinder on eBay for $11.50 and have ordered new abrasive wheels for it. From what I have read, the first thing that people do when they buy inexpensive new grinders is to replace the wheels, so I don’t think I am any worse off going this route than buying a new grinder. When this one wears out, I will have a much better idea of the type of grinder to look for, plus I will have become such a good turner that I can justify an expensive sharpening system.

My other purchase was a set of woodturning chisels from Penn State Industries. I got their six chisel set, item #LCSIXW. From what I can tell, they seem to be a very good set of tools, at least for learning with.

My current project is building a sharpening jig. I will try to post more pictures when I get it closer to done.

That is it for this time.

Have a great day!
Roger

Well it has been quite a while since I last posted to my Maranda blog. Progress has been slow, but there has been progress. The main thing that got in the way was building a new dresser for my wife and myself. The one we had been using was bought at an unfinished furniture store well over thirty years ago. So it was time.

The dresser is made from white oak that was cut on my father’s farm in Minnesota. He has been cutting saw logs and having them made into lumber, then he stores them in some of his sheds to dry. So to give you a peek at my handiwork, here is a picture of the finished product.

Now on to the topic that you are really interested in; the Maranda.

As I mentioned on the Maranda list, I am getting ready to build my wing spars. Since I am building the 14-H model, the wings use built up spars. Built up spars require some type of table or jig on which to build them and since each one is about sixteen feet long, the kitchen table will not work. Therefore, I decided to build a special table just for making the spars on.

With a long time of consideration I came up with a way that I could build the table out of one sheet of plywood. Well, there are a few additional scraps of wood that will be incorporated, but for the most part it all comes out of a single sheet of ply.

Starting with a nice, clear sheet of 1/2 inch plywood, I cut two strips that are 11 1/4 inch wide and 8 feet long. This width was chosen because that is the widest that my table saw will cut. One of these days I will invest in table extentions, but that is what I have now. After these two strips were cut, I cut the remaining plywood into strips 4 inches wide by 8 feet long. In this picture you can see the strips laid out to show how they will be assembled.

To assemble these, I first laid two four inch strips end to end and laid a third four inch strip on top of them so that the third strip was half on each. Then I used one inch long sheetrock screws to fasten them together. I drove some of the screws from each side of the assembly.

Next I turned my attention to the top. The top is comprised of the two 11 1/4 inch strips fastened end to end. To fasten them together I scrounged around in my garage and came up with a piece of left over half inch plywood about six inches wide and two feet long. I marked a line along the center in both directions on both sides. Then I marked the center line of the long strips on the bottom for a bit over a foot on each one. These lines are used for alligning all of the parts. Then, with the top pieces laying bottom up on the floor, I screwed the small plywood board to them. Now there are only three pieces.

The next step was to set the four inch strips upright, place the top on them, and screw them together. I made sure the uprights were snug against the center patch, then measured the distance to the edge of the top. I addes 1/2 inch to the measurement and marked it on the top to show where to place the screws. This works for the four feet on each side of the center of the table.

The next step will be to get some more material to build some legs and a bit of reinforcement for the finished assembly. However, even at this stage it is quite solid and stiff.

That is how things are going at this point. Hopefully it won’t take so long to post my next update.

This is my first post of Oracle DBA related stuff. Maybe I can remember to update this as I work on the databases and run into things that I think are interesting.

What I am working on this week is duplicating our development database to a testing environment. This has been done before several times by me, but for some reason it is giving me fits this time. Due to the size of the database the duplication takes about five hours to reach the point where it fails. That gets all the data files copied from the RMAN backup on the dev system to their proper location on the test system. This part seems to work quite nicely. The thing that is giving the problem is that at the point where RMAN is supposed to shut down the database and restart it, I get an error about the database not being available.

I printed out chapter 13, “Creating and Updating Duplicate Databases with RMAN”, from the RMAN manual. I thought about printing the whole manual, but it is almost 500 pages. I also got out Volume II – Student Guide from the Oracle Database 10g: Backup and Recovery class that I took last summer. I am referencing the “Solutions for Practice 6-1: Create a Duplicate Database” section in the class manual. By combining the instructions from each of these documents and filling in the holes from experience and Googling, I have been following this procedure:

1) System preparation and software installation:

The test database server is a Sun V210 with a newly installed copy of Solaris 9/05 that has had the patches 112874-33, 113225-08 and 118335-08 installed. I followed the Oracle 10g installation procedure laid out by Roger Schrag in his paper “Installing and Configuring Oracle Database 10g on the Solaris Platform” as far as installing 10.1.0.2.0 was concerned. We are running the bare first release of Oracle 10g and part of the function of the duplicated database will be to test upgrading to 10.1.0.4.0 and 10.1.0.5.0, but that will come later, so we just want the bare 10.1.0.2.0 Oracle software installed for now.

The patches that I installed on the system are to allow X fowarding across ssh and to implement the daylight savings time changes that went into effect last year.

The database was installed with the Oracle Universal Installer. I did not install any example or sample databases during the installation. Once the database is installed and running, the next step is to shut it down.

In order to create the duplicate database the source database server needs to be accessible from the test server. In my case I needed ports 22 and 1521 open both ways.

I am using RMAN to do daily full backups of the source database. These files are compressed and stored on a disk array that is mounted vis NFS across a direct (point to point) gigabit network connection. I duplicated this connection on the test system and mounted the drive with an identical mount point name. The duplicate database comes from the backup files, therefore I am (initially anyway) wanting to duplicate the database from development the way it was yesterday at this time. This will be important to keep in mind later on.

2) Setting up for duplication:

Information:
The ORACLE_SID will be “test”.
The password for the sys and system users will be “cafin8ed”

Task 1: Create an Oracle Password File for the New Database
This is fairly straight forward, but there were a couple of gaps in the information that I had. The password that you use is the password that you will use when you login as the sys user. Therefore, if you want to be able to use the command “sqlplus sys/charles as sysdba” the password that you should use with the orapwd command would be “charles”. The other thing that is not obvious is that the ORACLE_SID is case sensitive. If the SID of your new database is going to be “test”, then don’t set it some places as “TEST” and others as “test”. This is also used in naming your password file.

I like to use the password from my source database, just to keep things the same until after the duplication is completed. That would make the passwords on the source and target databases the same.

So, the first step is to go to $ORACLE_HOME/dbs and make sure that there is not a password file in there that has the same name as the password file that you intend to use. If you used your new SID during the installation of the database software there probably will be. Either remove the file or rename it. I like to rename it until I am sure that the new one is properly in place.

When that is clear we can create the new password file with the command:
orapwd file=orapwtest password=cafin8ed entries=10

Task 2: Establish Oracle Net Connectivity to the New Database
What this means is that you need to get your tnsnames.ora file setup correctly to allow you to connect to both your source database and to your new database. I simply copy the tnsnames.ora file from my source database to the new database, then edit it to add the entry for my new database.

The tnsnames.ora file from my source database:
# tnsnames.ora Network Configuration File: /u01/app/oracle/product/10.1.0/src/network/admin/tnsnames.ora
# Generated by Oracle configuration tools.

SRC =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = devdbserver)(PORT = 1521))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = src)
)
)

PROD =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = proddbserver)(PORT = 1521))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = prod)
)
)

EXTPROC_CONNECTION_DATA =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC))
)
(CONNECT_DATA =
(SID = PLSExtProc)
(PRESENTATION = RO)
)
)

The tnsnames.ora file from my new database:
# tnsnames.ora Network Configuration File: /s01/app/oracle/product/10g/test/network/admin/tnsnames.ora
# Generated by Oracle configuration tools.

TEST =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = newdbserver)(PORT = 1521))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = test)
)
)

SRC =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = devdbserver)(PORT = 1521))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = src)
)
)

EXTPROC_CONNECTION_DATA =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC))
)
(CONNECT_DATA =
(SID = PLSExtProc)
(PRESENTATION = RO)
)
)

Since these are different servers you need to be sure that the server names used in the tnsnames.ora files are known to both systems. I just updated my /etc/hosts files on each system. I also added an entry to my source database tnsnames.ora file for the test database, just in case I needed to access it from the source system at some time.

When these files are in place and the new instance is up you should be able to use tnsping to verify that you can access each of the instances. If you have problems, they will need to be corrected before continuing.

Task 3: Create an Initialization Parameter File for the New Database.
This parameter file is to get you going. It can be modified once the new database is up and running. The way to create this is to get a copy of the parameters that are currently being used on the source database. Therefore, logon to the source database server and decide where you want to put the initialization parameter file. Then create the file from the spfile (if you are using an spfile – most people will be) or copy the current pfile.

The command to create the pfile from the spfile is fairly simple. In my case I did the following:
SQL> create pfile=’/export/home/oracle/initTEST.ora’ from spfile;

Then you need to copy thie initTEST.ora file from the source system to the new database server. I just leave the file in my Oracle user’s home directory.

Now the “hard” part! The initTEST.ora file needs to be edited to prepare it for the duplication process. Here is what my init file looks like before editing:

SRC.__db_cache_size=4412407808
SRC.__java_pool_size=16777216
SRC.__large_pool_size=16777216
SRC.__shared_pool_size=1107296256
*._db_block_hash_latches=10240
*._recyclebin=false
*.compatible=’10.1.0.0.0′
*.control_files=’/u02/oradata/src/control.001.dbf’,’/u02/oradata/src/control.002.dbf’,’/u02/oradata/src/control.003.dbf’
*.db_block_size=8192
*.db_cache_size=8M
*.db_domain=’bucklehq.com’
*.db_file_multiblock_read_count=16
*.db_name=’src’
*.db_recovery_file_dest=’/u01/oradata/logs’
*.db_recovery_file_dest_size=59055800320
*.db_writer_processes=2
*.fast_start_mttr_target=600
*.filesystemio_options=’SETALL’
*.java_pool_size=0
*.job_queue_processes=250
*.large_pool_size=0
*.log_buffer=100000000
*.open_cursors=10000
*.pga_aggregate_target=650M
*.processes=250
*.remote_login_passwordfile=’EXCLUSIVE’
*.sga_max_size=5400M
*.sga_target=5670699008
*.shared_pool_size=104857600
*.timed_statistics=TRUE
*.trace_enabled=false
*.undo_management=’AUTO’
*.undo_retention=1800
*.undo_tablespace=’SRC_UNDO’

There are a number of changes that I made, so I will step through them one at a time and try to explain why I did each one.

The first thing to do is get rid of any entry that starts with the old instance name. In my case that is anything that starts with “SRC.”. Just delete these lines.

Next, the new instance needs to know where it should put the control files and what to call them. Edit the line that begins “*.control_files=” to identify where your new control files should be located.

Be sure to change the db_name from the source database, “src”, to the new database, “test”. Be sure to be careful about the case!

Do not change the db_block_size value! If it is set leave it’s value. If it is not set, then don’t set it.

I wanted to change the location where the archive log files were stored, so I changed the value of the line that starts “*.db_recovery_file_dest=”.

I also wanted to specify where my dump files and audit files went, so I added lines to set the audit_file_dest, background_dump_dest, core_dump_dest and user_dump_dest locations. If you already have these lines you may want to change the values to meet your needs.

The last thing to do is to tell the system where to put the datafiles. In my case I was changing the path to the various files. At other times I have gone from a source system with the files spread across multiple small disks and moved to a system that had a single partition on a disk array. Use the DB_FILE_NAME_CONVERT parameter to change the file names at this point. Use the LOG_FILE_NAME_CONVERT to change where the log files are located. In each case the values given to the *_CONVERT parameter need to come in pairs: ,. You can even change the location of each particular file by putting a value pair for each file that will be created on the new system. I haven’t had a need to become quite so fancy yet.

So after all the editing, here is my finished initTEST.ora file:

*._db_block_hash_latches=10240
*._recyclebin=false
*.compatible=’10.1.0.2.0′
*.control_files=’/s01/oradata/test/control01.ctl’,’/s01/oradata/test/control02.ctl’,’/s01/oradata/test/control03.ctl’
*.db_block_size=8192
*.db_cache_size=8M
*.db_domain=’bucklehq.com’
*.db_file_multiblock_read_count=16
*.db_name=’test’
*.db_recovery_file_dest=’/s01/logs’
*.db_recovery_file_dest_size=55G
*.db_writer_processes=2
*.fast_start_mttr_target=600
*.filesystemio_options=’SETALL’
*.java_pool_size=0
*.job_queue_processes=250
*.large_pool_size=0
*.log_buffer=100000000
*.open_cursors=10000
*.pga_aggregate_target=650M
*.processes=250
*.remote_login_passwordfile=’EXCLUSIVE’
*.sga_max_size=5400M
*.sga_target=5670699008
*.shared_pool_size=104857600
*.timed_statistics=TRUE
*.trace_enabled=false
*.undo_management=’AUTO’
*.undo_retention=1800
*.undo_tablespace=’WSD_UNDO’
audit_file_dest=’/s01/logs/adump’
background_dump_dest=’/s01/logs/bdump’
core_dump_dest=’/s01/logs/cdump’
user_dump_dest=’/s01/logs/udump’
DB_FILE_NAME_CONVERT=’/u01/oradata/src’,’/s01/oradata/test’,’/u02/oradata/src’,’/s02/oradata/test’
LOG_FILE_NAME_CONVERT=’/u03/oradata/src’,’/s01/oradata/test’

Task 4: Start the New Database
There are two documented ways to do this. Either way should work, but I have had the best luck with the method that is taught in the Backup and Recovery class.

Be sure your SID is set correctly! This is where I think I have gotten into trouble with case sensitivity. So for our example do the following:
$ ORACLE_SID=test
$ export ORACLE_SID
$ echo $ORACLE_SID

The response should be “test”.

Now login to sqlplus as the sysdba:
$ sqlplus / as sysdba

You should now be connected to an idle instance.

Startup the instance using the initTEST.ora file that you have just gotten done editing:
SQL> startup nomount pfile=’/export/home/oracle/initTEST.ora’

Then create a server parameter file from the pfile:
SQL> create spfile from pfile=’/export/home/oracle/initTEST.ora’;

When you are told that the file has been created you can exit from sqlplus.

Task 5: Mount or Open the Source Database

In my case the source database is always in use, so it is always open and mounted. However, it is a good idea to verify that you are able to access the source database from the new system, just to be sure. Therefore, do the following:
$sqlplus sys/cafin8ed@src as sysdba
SQL> select open_mode from v$database;

You should see that the database has an open mode of “READ WRITE”.

3) Do the duplication!

If everything has gone ok so far, you should be ready to kick off the actual duplication commands. I like to script these, but they can be done from the command line just as easily. The main reason that I script mine is that I like to kick off a duplication just before I go home and let it run in the background. So I set up a shell script that calls RMAN with CMDFILE and LOG parameters. Then I can come in the next morning and see what happened by looking at the log file.

My duplication script looks like this:
#!/bin/ksh

#-> set -x

ORACLE_SID=test
export ORACLE_SID

ORACLE_BASE=/s01/app/oracle
export ORACLE_BASE

ORACLE_HOME=/s01/app/oracle/product/10g/test
export ORACLE_HOME

PATH=$ORACLE_HOME/bin:$ORACLE_HOME/ccr/bin:/usr/bin:/opt/sfw/bin:/usr/local/bin:/usr/ucb:/etc:.
export PATH

NLS_LANG=AMERICAN_AMERICA.UTF8
export NLS_LANG

NLS_DATE_FORMAT=YYYY Mon DD HH24:MI:SS
export NLS_DATE_FORMAT

RMAN=/s01/app/oracle/product/10g/test/bin/rman
COMMANDFILE=/export/home/oracle/RMAN/copy.scrpt
LOGFILE=/export/home/oracle/RMAN/copy.log
RUNCHECKFILE=/export/home/oracle/RMAN/copy.out

echo “Running RMAN duplication of SRC to TEST at `date`” >$RUNCHECKFILE
echo >>$RUNCHECKFILE

$RMAN CMDFILE = $COMMANDFILE LOG = $LOGFILE

echo >>$RUNCHECKFILE
echo “RMAN duplication of SRC to TEST completed at `date`” >>$RUNCHECKFILE
echo >>$RUNCHECKFILE
echo “Please check the setup log \”/export/home/oracle/RMAN/copy.log\” for details.” >>$RUNCHECKFILE

My RMAN script file that has the commands to actually do the duplication looks like this:

connect auxiliary sys/o4guk8x
connect target sys/o4guk8x@src

run
{
allocate auxiliary channel aux1 device type disk;
allocate auxiliary channel aux2 device type disk;
allocate auxiliary channel aux3 device type disk;
duplicate target database to test until time ‘sysdate – 1’;
}

The reason that I allocated three channels was because the Oracle documentation says “If the backups reside on disk, then the more channels you allocate, the faster the duplication will be.” I have not experienced any speedup in the duplication process in going from one channel to three channels. However, I am going to try adding more channels to the backup as well to see if that will help. I suspect that increasing the number of channels will only work up to the number of channels that were used to create the backup. This is something that I will be checking out later. Right now it is taking about five hours to do a duplication in my environment.

What I do to start the duplication is:
$ nohup ./copy.sh &
$ tail -f copy.log

and just sit and watch it do its stuff.

That is it for now. I hope you enjoyed this and found it useful.

Well, even though it has been quite a while since my last post, things have been getting done on the plane.

The most progress has been made on the engine. I had two core engines that were both “stuck” and was trying to get at least one of them “un-stuck” to use as the base for the finished engine. Unfortunately, neither engine would give way on their own, so we had to resort to taking them apart the best we could while they were still frozen.

The engine that I had been working on the most is the one that was on my engine cart. I had pulled the valve covers, the rocker arms and push rods. Then I pulled the spark plugs and filled the cylinders with kerosene and let them soak for about three weeks. Still not luck with getting it to turn!

Well, a bunch of guys (about six) from church wanted to do something as a group project, so on Saturday, March 1st, they came over and we proceeded to tear down both engines. Between 9:00 am and 2:00 pm we got both engines completely torn down. Now I am going through the parts and selecting the best ones to build the new engine from.

When we pulled one head on the RL core about two cups of very fine sand poured out of one cylinder! No wonder we couldn’t get that engine to turn.

There was essentially no oil in either engine, so there was very limited mess during the disassembly process.

Out of 12 cylinder/piston assemblies there were 7 that were not frozen. The others are VERY rusted. I am hoping to clean up all 7 of the good cylinders and have them bored.

The crank in the RL core appears to be factory nitrided! This is good news if it is solid otherwise. It will still need to be magnafluxed to ensure that there are no cracks in it, then it will need to be evaluated to see if it needs to be reground. I am a bit concerned about this crank as some of the connecting rod end bearings look like they may have spun. This tag on this engine said it came from a car with a Power Glide transmission, but the bell housing was for a manual transmission and the pressure plate was still in place on the engine, so I don’t think that the crank had the 4 degree offset.

The cam shafts all looked good, but I will need to trade up to an OT-10 to get the proper performance.

I cut down one of the bell housings to make the front end cover for the new engine. I used a power scoll saw with a hacksaw blade in it for most of the cutting. This worked quite well, but the blade wore out before the job was completely done. I finished up with my hand hack saw. Now I need to do a bunch of filing to finish the edges the way I want them.

The cylinders are pretty rusty on the outside. I have started cleaning them up with a wire brush, but it will take a lot of brushing to get them ready to paint this way. I am going to make use of some of the stuck cylinders to experiment on ways to clean them up. I am thinking of trying sand blasting, naval jelly and electrolysis. And I might need to get a bench grinder and fit it with a wire wheel.

On the fuselage end of things I have been pulling staples from the ribs and looking for wood. Lowes in Lincoln has some very nice Douglas Fir but the longest lenght is 12 feet, so that would require scarfing every longeron for the fuselage and most of the really long pieces for the wing spars. I checked out both Menards and Builders Warehouse and came up with nothing. However, I just happened to stop at Foster Lumber and they can order anything that I want, so I am going to have them get me a quote for Douglas Fir in 16 foot lengths and Sitka Spruce as well. I can probably even have them deliver it to my house.

That about wraps it up for now. I don’t have any picture updates today.

Have a good one!