Disk Usage in Dashboard

[ganesh.usi.gk]

Let me explain with below examples

Assume I have only one database in the host.

Now my database size is 1GB. (Disk Usage of database)
And host size is 25 GB. (Total disk usage shown for cluster in dashboard)

Now just am loading the data which is just 100 MB.

So my database size is increasing up to 1100 MB ( Just adding 100 MB with existing 1GB)

But my host size is increasing up to 29 GB .

This is what I don’t understand . Instead of 100 MB why 3GB has been increased. I can understand some extra disk space will occupy for logs ,query and etc db activities.

But for just 100MB data loading ,how increasing some amount of GB level in cluster disk usage level?

If this is the case , How DISK USAGE (Host size) will increase in future when we go with actual data?

[adam]

I think this post should help explain some of what is going on:

Disk usage of database logs

Thanks for that info - the write workload is certainly aggressive during the day for us (100K+ records/sec at times), which would explain why we have 5-6 pre-allocated log files even for an empty db (with 8x2 partitions per leaf, because of redundant mode). [ec2-user@ip-172-31-16-188 logs]$ du trading_prod_* | cut -f1-3 -d_ | sort | uniq -c 6 262144 trading_prod_0 6 262144 trading_prod_1 6 262144 trading_prod_10 5 262144 trading_prod_11 5 262144 trading_prod_12 6 262144 tradin…

Is the data in columnstore or rowstore tables?

[ganesh.usi.gk]

Thanks Adam,

It is helping me to understand how disk usage is calculated.

But occupying 40GB for 600 MB Table data is some thing wrong or is this behavior of singlestore db architecture.

( i have executed manually snapshot databse command and my snapshot size is 2 )