Consultation with disk space

Ops and Admin
1

Good fellows I have a question that I would like to solve: We have a Tier M30 cluster with 130 GB storege with storage scaling activated.
Well currently we have lowered the disk space to about 114.0Gb in the next few days we will continue lowering.
Is it advisable to manually lower the GB of disk space? Will there be any kind of stop or affect the production startup?
I also do not understand how my collections in total occupy 109.594GB and instead in the monitoring it continues putting Disk Usage 114.0 GB, where can I free space? Can it be configured to scale down as it does automatically when it needs to expand?

Thank you very much are several questions but it is a topic that has me a little disorienting, thanks once more

2

Hi @Juan_Jose_Garcia_Gonzalez,
I will try to help you!

Doing a very simple calculation, 90% of 130Gb is 117Gb, so the next time when will scale the cluster storage.

From the documentation:

These three parameters should answer your question:

  • size
  • storageSize
  • totalSize

I don’t seem to have found any warnings inherent in manually reducing storage, but I personally wouldn’t do it.

Regards

3

Hi and welcome to the forums @Juan_Jose_Garcia_Gonzalez

As items are removed from the collection the blocks are simply marked as free for reuse for future document updates or inserts. To recover any free space to the OS you would need to compact the collection.

Directly connect to each secondary in turn and compact the collection(s) then use the “Test Failover” to switch Primary to another member and repeat the compact on this member.

But @Fabio_Ramohitaj’s advice remains valid, you would be close to storage autoscaling threshold. Maybe the cost/benefit works out for you though!

1 Like