Looking for an alternative database to MS SQL to process a text file.
-
@magicmarker
I see 2 conditions for that to work:- Never process the file until it is completely written/received
- The bar code software must not lock the file exclusively so all branches can use it (possibly) at the same time
-
@scottalanmiller said in Looking for an alternative database to MS SQL to process a text file.:
Since you are looking to learn here, as well as accomplish a goal, it is worth noting that a Relational Database like SQL Server, MariaDB, Oracle, Informix, DB2, PostgreSQL, MySQL, etc. is not the right kind of database for this work. This does NOT mean that you should change now, you've done the work, and learning something completely new would be silly. And at this scale, anything that does the job is fine.
Relational Databases are designed around doing complicated relational integrity. They are "big and heavy" systems because of this. What you are doing is SO basic. What you would want, in an ideal world, is something way simpler from the NoSQL database family (NoSQL simply means that it is something other than a relational database, it can be ANYTHING else.)
A really simple key-value store sounds like it would meet your needs better. Way simpler, less to learn, less to know, less to do. REDIS would be the obvious choice due to market penetration, popularity and ease of use.
Again, not saying to switch. Saying that when it comes to learning about this, this is a better way to have approached it. REDIS is purpose built for this kind of need.
Thanks for posting this Scott. Very interesting. I quickly googled REDIS, but it's over my head. Any other good resources you can provide to help me learn more about this?
-
@dave_c said in Looking for an alternative database to MS SQL to process a text file.:
@magicmarker
I see 2 conditions for that to work:- Never process the file until it is completely written/received
- The bar code software must not lock the file exclusively so all branches can use it (possibly) at the same time
The BarTender label software does not lock the text file. I have confirmed that with support today. BarTender just does read-only against the text file when it queries the file. The only problem that I could see happening is someone attempting to lookup data (print a label) on the text file the split second it takes to overwrite the file as Scott mentioned. BarTender does not read the file until you query the text file for a lookup. It think it would be very rare that that a lookup happens the exact time the file saves/overwrites, but it could happen.
-
@magicmarker said in Looking for an alternative database to MS SQL to process a text file.:
@dave_c said in Looking for an alternative database to MS SQL to process a text file.:
@magicmarker
I see 2 conditions for that to work:- Never process the file until it is completely written/received
- The bar code software must not lock the file exclusively so all branches can use it (possibly) at the same time
The BarTender label software does not lock the text file. I have confirmed that with support today. BarTender just does read-only against the text file when it queries the file. The only problem that I could see happening is someone attempting to lookup data (print a label) on the text file the split second it takes to overwrite the file as Scott mentioned. BarTender does not read the file until you query the text file for a lookup. It think it would be very rare that that a lookup happens the exact time the file saves/overwrites, but it could happen.
Should be locked during the write. Probably not a concern.
-
Hmm, I work with these things quite often.
I suggest changing the name of the text file. That will be an atomic operation. I suggest doing it after the file has been written to disk. If you try rename operation while the file is written to, it will fail. After it has been renamed you have all the time in the world.
But first you want to check the integrity of the text file before import. How can this be done? Also you want to have a log file created so you can trace things back. You want to make damn sure your stuff is working and have traceability. We are talking about hazardous stuff, liability etc.
Then you can put the data in the database. I suggest either archiving the text files and/or writing timestamped info in the db. You could have one table with all the hazardous info ever received and another one with just the current one under production.
To do this you need some simple programming. Any script language will do. For instance php in cli mode, python or whatever you feel comfortable with. If you don't feel comfortable, hire a pro. To me it sounds like a days work, maybe two, all in all.
Oh, I forgot that you should look into having your importing utility send an email if something goes wrong. Or monitor it some other way. Software like this have a tendency to keep working unnoticed and then no one notices when they stop working for some reason.
PS. SQL Express has a 10GB data size limit, not 1GB. Otherwise if you are running windows I suggest MySQL as an alternative. They have a nice Workbench software that you can use to administer the db. https://www.mysql.com/products/workbench/
This is where you find the free (community) version of MySQL:
https://www.mysql.com/products/community/ -
@Pete-S said in Looking for an alternative database to MS SQL to process a text file.:
Hmm, I work with these things quite often.
I suggest changing the name of the text file. That will be an atomic operation. I suggest doing it after the file has been written to disk. If you try rename operation while the file is written to, it will fail. After it has been renamed you have all the time in the world.
So the flow would look like this then? Receive text file > Rename text file > copy text file to file server shared folder > rename original text file > rename new copy to original name > archive old text file.
Did I understand that right?
-
@Pete-S said in Looking for an alternative database to MS SQL to process a text file.:
PS. SQL Express has a 10GB data size limit, not 1GB. Otherwise if you are running windows I suggest MySQL as an alternative. They have a nice Workbench software that you can use to administer the db. https://www.mysql.com/products/workbench/
I said my databases are under 1GB, not that MS SQL Express has the 1GB size limit.
-
@magicmarker said in Looking for an alternative database to MS SQL to process a text file.:
@Pete-S said in Looking for an alternative database to MS SQL to process a text file.:
PS. SQL Express has a 10GB data size limit, not 1GB. Otherwise if you are running windows I suggest MySQL as an alternative. They have a nice Workbench software that you can use to administer the db. https://www.mysql.com/products/workbench/
I said my databases are under 1GB, not that MS SQL Express has the 1GB size limit.
And it was stated in like post 6 that Express was 10GB. :thumbs_up:
-
@magicmarker said in Looking for an alternative database to MS SQL to process a text file.:
@Pete-S said in Looking for an alternative database to MS SQL to process a text file.:
Hmm, I work with these things quite often.
I suggest changing the name of the text file. That will be an atomic operation. I suggest doing it after the file has been written to disk. If you try rename operation while the file is written to, it will fail. After it has been renamed you have all the time in the world.
So the flow would look like this then? Receive text file > Rename text file > copy text file to file server shared folder > rename original text file > rename new copy to original name > archive old text file.
Did I understand that right?
I don't know. I assumed the text file with updated hazmat info have the same name each time you get it and that it ends up in the same location.
So then it is:
1 Some process writes a text file somewhere, say info.txt
2 You detect that there is a info.txt and you rename it to something like info_000232.txt
3 You ingest the renamed text file into the db
4 Archive the text file by moving it to a folder with archived text files. Maybe compress it as well.As soon as you have renamed the file where it is (step 2), you are ready to receive a new info.txt without having to deal with partially overwritten files. This works well also over ftp or whatever you use.
When you rename, use a sequential number or put the time and date in the file name so it becomes unique and you can find it.
-
@JaredBusch said in Looking for an alternative database to MS SQL to process a text file.:
@magicmarker said in Looking for an alternative database to MS SQL to process a text file.:
@Pete-S said in Looking for an alternative database to MS SQL to process a text file.:
PS. SQL Express has a 10GB data size limit, not 1GB. Otherwise if you are running windows I suggest MySQL as an alternative. They have a nice Workbench software that you can use to administer the db. https://www.mysql.com/products/workbench/
I said my databases are under 1GB, not that MS SQL Express has the 1GB size limit.
And it was stated in like post 6 that Express was 10GB. :thumbs_up:
My bad! :loudly_crying_face:
-
Why dont they create one giant hazardous label to fit all the products, and make it generic as possible.
Like this product contains hazardous chemicals.
it is not like someone actual reads that small fine print.
Another problem solved...
