Reputation
2,702
Top tag
Next privilege 3,000 Rep.
Cast close & reopen votes
Badges
1 6 14
Newest
 Scholar
Impact
~335k people reached

2d
comment Does table size matter?
The index depth depends on data diversity. But AGAIN: IT should not matter. If you have 5 levels now, you likely add only one. And that has only one page, or something like that, so the overhead LIKELY is very low. As long as you do not touch more than last day (regularly).
2d
answered Does table size matter?
2d
comment Move/ Migrate SQL Server 2012 Virtual machine (Win 2012) to new Physical host
You are confused. Moving / migrating VMs between hosts is trivial (and not a dba question - this is standard in any larger virtualization environment) - but then you go on a tangent talking about HA, which is totally unrelated to the question in the header.
Apr
24
comment The definitive fastest radius search query with Postgis 2.0
No, not done GIS level stuff for years and SQL Server (which i Use) has spatial indices for this. This is geometry optimization 101, though (even in 3d). Find the extremes on X and Axis (center point + and - radius) and make this the 4 conditions of the WHERE clause for the first filter.
Apr
24
comment The definitive fastest radius search query with Postgis 2.0
Reality check: FIRST make a BOX filter for an enveloping box for the circle - that is a lot faster and requires no distance calculation. Then post-filter to get rid of the little rest. This is how any sensible GIS level function will work. Bounding box checks are fast (4 fixed conditions), distance calculations are always expensive.
Apr
21
comment If replication is 600 seconds behind master will INSERT or UPDATE queries including `NOW()` be 600s different?
Interesting. I personally think that would not work - not under all circumstances. As I said - MS does actually replicate by tracking changed rows and then generating statements from that. But then, they do not ONLY pure Master/Slave... a Slave may well be filtered (not have all data) so an update statement may do different on a slave. I look forward to read the answers here.
Apr
21
comment If replication is 600 seconds behind master will INSERT or UPDATE queries including `NOW()` be 600s different?
Does MySql really work this way? On SQL Server, replication would not replicate the original statement. The Insert on the slave would be generated from the inserted row. I can see risk with data being inconsistent if replication works as you assume it does, mostly when you do update statements or inserts that select from other data.
Apr
21
answered MySQL - Like with index
Apr
13
answered Replicate only specific records of a table
Apr
11
comment Basic PostgreSQL 9.x Performance Optimization Strategies for Reads/Analysis
SSD. Any decent mid range (SATA) SSD will provide 100 times the IOPS budget of a SAS disc. If you do not have enough RAM, you need IOPS-
Apr
11
comment Basic PostgreSQL 9.x Performance Optimization Strategies for Reads/Analysis
Disc layout? Why use something as cost and performance inefficient as SAS discs?
Apr
10
answered Primary key which is foreign key in an relationship is specified NOT NULL. Why?
Apr
8
asked Matching table (with sub-tables) to another table(with sub-tables) efficiently
Apr
8
awarded  Scholar
Apr
8
accepted Custom ID Numbering - fast way on insert?
Apr
8
comment Custom ID Numbering - fast way on insert?
THen the bucket is reset for the next data set. I really look for something that works dynamically based on the existing data.
Apr
8
comment Custom ID Numbering - fast way on insert?
Gets complex. Espcially because I need this about 20 times for different data(so that is a LOT of sequences, also regularly resetting)
Apr
8
comment Custom ID Numbering - fast way on insert?
Yes, there are. 4096. Buckets are basically "storage bins" for data (and correspond to a table partition each).
Apr
8
asked Custom ID Numbering - fast way on insert?
Apr
8
comment Do you need SQL Server AAG when using SQL Server FCI/RAID10?
Actually no, FCI Only hits your requirements if you can live with the database going down due to data corruption. There is a LOT that can go wrong with SAN level replication (such as the SAN happily replicating a corrupted database thanks to a controller malfunction or server RAM error). AOAG are explicitly there to avoid the single point of failure that the database file is.