Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations strongm on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Backend Bloat

Status
Not open for further replies.

lameid

Programmer
Jan 31, 2001
4,207
US
I have a backend database that has been observed bloated it the 390 MB range and on compact is under 7 MB.

I am well aware of the difference in record locking and not and record locking is turned on.

Interestingly the file after compact was 6600 KB. Then I ran a particular process that adds records (<100) amongst 4 tables. After this process it shot up to 113,764 KB. Then compact and down to 6680 KB. Next I stepped through the same code and it grew to 6948 KB. A subsequent compact and down to 6720 KB. This latter test is in line with what I would expect, a few extra hundred k for record locking. Now I have been unable to repeat the first crazy growth. Any thoughts on what causes this sort of thing or may be intermittent?

I suspect something about code but rather than a general go implement best practices direction, I was hoping someone had some specific thoughts on what to fix first.

Typically this is a multi-user system but my test was done with a copy of the BE / FE locally and attached locally.

As a side note we recently ran across a similar issue for Front ends when a recordset is not closed properly.... Access recompiles the code multiple times, bloating the file. A reason I tend to look first at legacy code.
 
It is repeatable... Compact the source copy backend get the 6660 KB file (oops I was off 60K when I typed) and then get the huge growth...

Now the weird part compact the 6660 KB file... get a 6652 KB file... run the same process again... 2 out of 3 times the database bloat is hundreds of KB as expected.
 
After checking other systems with the same code, this seems to be anomalous rather than the rule. I am chalking it up to sometimes bad things happen in the computing world. I still suspect the compiler and code not being less than ideal.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top