Speed It Up on LARGE Databases
I created a new project. DB has 2004 tables. I only care about ONE db schema - "pricemanager", so, about 20 tables, of which I care to compare just 11. It takes nearly five full minutes just to get a tables and views page I can choose from.
WHY, can I not simply "FILTER BEFORE RETRIEVAL"? I know what I want, and it isn't 2000+ tables and 5000+indexes, et al.
I JUST upgraded to 13.4.5.6953 in hopes of some speed improvement
- all hopes were dashed when it ground on for the usual 5 mins.
need to eliminate the 4 typical "who" columns for each of my 11 tables bearing these columns. This is, arguably, one of the WORST experiences a Product has to offer... It is SO slow!
It is plain ridiculous to have to wait another five to ten minutes (NO EXAGGERATION) for every table I need to simply omit columns from... And while here, it is also RIDICULOUS that to edit my project to include another table, I have to wait ANOTHER FIVE MINUTES for "it" to go back to the DBs I already went to, just to collect the SAME objects, in the SAME state... GIVE US A "DO NOT RE-READ SCHEMA" OPTION! (Both while we have a project open and need to edit, OR when we first open SQDC - allow us to say, "JUST GET MY PROJECT'S CHECKED TABLES OBJECT DEFINITIONS" (getting 11 instead of 2004, now that's a speed-up waiting to happen...)
I cannot attest to this Product being a "productivity enhancer" - simply because I waste tens of minutes while unnecessary schema reads are repeatedly done, and "it" goes off into la-la land simply when I exclude 4 columns from the a table compare! Shocking it should be this bad at Version 13!
And, when I have the GRID of results on show, GIVE ME A RIGHT-CLICK TO EXCLUDE A COLUMN, there and then, while the grid is on view - and "REFRESH JUST THIS TABLE" taking into account newly excluded columns... SO. MUCH. FASTER. (Mark this "un-map" in the Project, too, PLEASE)
And this is on a MacBook Pro, with i7 processor and 16GB RAM, and SSD, so it's no slouch - and the servers are via wired 10GB Ethernet, about 150ft away, so no network latency.
ANY pointers RG can give to make this a usable tool would be greatly appreciated!
-
Anonymous commented
Same here - offer an option to
a) just read and list out all tables and columns, then let the user do the mapping manually. Do not try to default auto-map!
b) refresh the list when the user explicitly requests it (refresh button)I have 5000 tables with a total of 1.1 Mio columns; and all I want to do is compare data in one single table that has 6 columns. I dont care about any of the other tables!
And please don't ask me now if it wouldn't make sense to rearrange the tables and break some of them out into a different DB...