Hello,

On 3/1/2006 7:26 PM, Christopher Mills wrote:
Arno,

Thank you for taking the time to review my long post, and for steering me on course. I checked the bug tracker page, which I just discovered. The bug was reported and fixed, so I will wait for a future release (bug 0000339). I will check the buglist next time, not just the user posts.

A few follow-up questions:

1) I didn't follow your blacklist comment at the beginning.

That referred not to a blacklist on my side or the lists, but rather to a problem my mail provider seems to have. They managed to get at least one of their outgoing mail servers onto a blacklist sf.net queries and so lots of my list mail was refused and I had to resend it and complain to my provider. Just to keep track of which message I tried to resend I addrd something to the top of it :-)

Other than length, have I violated any etiquette?

Not as far as I'm concerned. And, just to make this more clear, the length is in my opinion not a problem. It's merely an observation that shorter posts get more responses on many mailing lists, even if it takes quite a lengthy exchange to get all the facts you need to actually understand what is asked or reported.

2) Is there a repository of queries written by others?

The best one I know is in query.sql - seriously, I don't think so. There is something in there that I contributed, I think, but I suppose that many others work similar to the way I do: Create an SQL hack, tweak it until it does what we want, and unfortunately, at that point, it's definitely not portable, elegant or even readable an more, so we keep this in our installation and don't share it. (Might be something for the wiki, I think).

3) Is there a pictorial view of the data-model,

You might find something in the development manual, but personally, I never needed such an overview. Given the table names and the corresponding Bacula terminology, half the stuff I needed was already there, and the other half are the links between tables. These are usually an as far as I know not managed by database keys, but purely in the application logic. The naming helps a lot to understand these connections: For example the links to the Job table are done by a column called JobId which contains the calue of the primary key of the Job table, which is also called - you guessed it - JobId.

Once I got that far I could work my way through the data model as far as necessary.

or do you know of any good reverse engineering tools for mysql that will build me a nice graphical picture of the data model?

Don't know nothing about such stuff, and I suppose it would need lots of user knowledge anyway, because all the links are managed in Bacula, not the database backend.

I have a reasonable amount of experience in relational data modeling. Unfortunately, my experience is in a proprietary (albeit robust) relational database. Consequently, I am lukewarm on sql queries from a practical standpoint, and almost completely unfamiliar with any mysql tools that are out there. Any must-haves?

I always use mysql, the simple command-line shell. As a graphical management tool I prefer MySQL Administrator, freely available from the MySQL web site. But I prefer doing "real" work in a terminal window or similar, so YMMV.

Arno


Arno Lehmann wrote:

Once more, again blacklisted...

this gets interesting ;-)

And once more because I sent this to the wrong list...

resend, because my mail provider managed to get one of his mailers onto
a blacklist...

Hello,

you wrote a very nice mail - only I think many list-readers will not
bother to read that long text ;-)

On 2/28/2006 11:39 PM, Christopher Mills wrote:

I searched the archive for an answer to this question. I found it asked but never answered:

The bconsole command 'list jobs' gives me a list of runs from all jobs. I want to narrow it to all job runs from a particular job. The bconsole doc under "list" seems to imply that would be done with:

  list job=My-Job

This doesn't work (I am running mysql under Gentoo Linux).



The same here. Either the manual is wrong or this is a bug, I say.

I tried ...
Alas, my suspicion is this query just doesn't function.



By now, I suspect the same.

Since I believe that this is a very fundemental query, and since out of the corner of my eye while searching the archives for an answer I seem to have noticed questions about formulating sql queries against the Bacula database tables, I have a broader question in regard to what I should expect in diving into Bacula:

Should I expect that the Bacula development community is focusing on the more critical areas of this highly complex and powerful opensource project?



That's what it looks like, and even though I understand the question
behind what you wrote :-) I don't think this should remain unfixed.

Should I bite the bullet, and bone up on my sql skills on the assumption that these relatively benign (but from a practical perspective of a Bacula implimentation, important) queries may not be fully debugged and tested as of yet? Should I assume that I will "become one" with the product more quickly if I just accept this, and learn to write my own queries as needed?



Well, if you want to get more fluent in SQL this could be just the right
project :-)

I am really hoping the answer is "no." But I can understand if it is "yes," as I certainly haven't been the one contributing to this monumental project :) (Although, if it is "yes" I will confess to being a little disappointed as I was hoping query writing would only come up in future fine-tuning of my implimentation).



I'd even say that in many cases you don't need to write any SQL at all.

If this is the case, then I am just wondering where I should start. I haven't really tried to delve into the source although I have taken a peek at "query.sql." What is not clear to me is where all the queries forming the 'list' commands are hiding. Do I need to get my hands on some source to get started on examples of useful queries? Should the task be to simply tack new queries onto query.sql as I need them?



Yes and no.

The SQL behind the commands like list is indeed hiding in the source
code. You can, however, achieve what you want by using queries in
query.sql, which I'd recommend.

Parameter passing is done in a dialog there, but it's obviously much
easier to try out queries - no need to recompile, install and restart
the binaries, only saving the file and issueing the query command from
bconsole.

Or am I getting paranoid, all documented queries work great, and I am just being clueless, having missed the obvious (which wouldn't be the first time)? :)



Not this time, I'm afraid...

Thanks in advance for any help and advice.


-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting language that extends applications into web and mobile media. Attend the live webcast and join the prime developer group breaking into this new coding territory!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users




--
IT-Service Lehmann                    [EMAIL PROTECTED]
Arno Lehmann                  http://www.its-lehmann.de


-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting language
that extends applications into web and mobile media. Attend the live webcast
and join the prime developer group breaking into this new coding territory!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users

Reply via email to