Over the years, we’ve heard the question about the size in GiB of the ERP instances repeatedly. This question, quite logically, is coming to mind because of the fact that we charge hosting based on used GiB (Gibibyte).
With this post, we would like to provide as much information on the subject as possible.
First, let us mention the fact that these amounts ARE NOT only for the usage of the data storage disk. Rather, this is a measure of the load on different systems: disks, memory, traffic, backups, etc. But even if we are just talking about disk systems, for 1 TiB of user data we need approximately 20 TiB of disk arrays! Apart from that, we have memory load, networking and processing load, and reserved cloud resources.
In fact, we can call 1 GiB of user data “1 unit” and we can say that with that unit we measure everything. The reason for selecting such a unit is simplicity. The standard cloud systems (like Microsoft Azure, Amazon Web Services, etc.) have very complicated measuring systems for calculating the usage of cloud resources. We have put it all down to a simple unit for our client’s convenience. However, this should not be mistaken (as it happens quite often) with simple data storing on a disk. It is a complete load’s measurement.
What are the options for optimizing this usage?
It is important to note that once a certain usage is reached it CAN NOT be lowered. Theoretically, there is such a technical operation. However, it fragments the user data too much and *forever*. This means that one time lowering of the usage slows down the productivity *forever*. Therefore, we are not using this option at all.
At the same time, it is perfectly possible (with version 2020 and above) to free up internal space. Freeing up internal space means that the future growth of the instance would be much slower. Due to a number of technical details, complete stopping of the growth is impossible, but lowering it is completely feasible.
Note: Users’ ERP instances are often referred to as “databases” because this is the internal technical format of the main hosting. However, this is not completely accurate and, in the future, the internal format might be divided in different ways.
In a user’s ERP instance roughly half of the data is historical, such as:
- Track changes (and the old Document Versions)
- Audit logs
- Document printouts
Since version 20, we started creating the so-called „jobs“ which could be run automatically at night to delete old records (usually older than 3 years). This means that the automatic jobs free up internal space in the instances and slow down the volume’s growth.
In addition, in version 22, there is an option to start job execution manually, not only automatically.
More information on “Jobs” you can find here: https://docs.erp.net/tech/advanced/jobs/index.html and here: https://support.erp.bg/hc/en-us/articles/4410153850770