You are here: Running solutions in production > The Server > Disk space guidelines

Disk space guidelines for the server

MD Link's persistence database may consume a lot of disk space when running the server over long periods. If one disables the solution-level auto-purge options, or sets them to very high values, or is running MD Link on a machine with limited disk space, then one will need to pay attention to this. What follows is a rough disk space guideline for the MD Link server.

For a typical solution handling medium-sized HL7 messages, the amount of disk space required under the data folder, to hold both message data and log messages, can be approximated by the following formula:

<Number of messages per meg of disk space>

= 250 divided by <the number of components in the solution>

Equivalently:

<Number of megs per day>

= <number of components in solution> times <number of message per day> divided by 250

The number of components (i.e. events + resources + tasks) is typically the major determining factor. The size of those messages, incidentally, does not have a significant impact - at least not for messages in the typical size range for MD Link - 2000 characters or less.

Furthermore, if you have large input queues building up, then the additional amount of disk space require by each input queue is approximately 1 meg for every 5000 input queue elements. This is much smaller because internally, the queue structures don’t contain the message payloads – they only contain references to the message payloads, which are stored in the part of the database addressed by the message data guideline, above.