B DSR VM Disk Requirements

This section provides guidance on the disk requirements for the OCDSR VMs. Characterizing disk requirements can be tricky since there are many variables that can affect disk usage, such as how many reports are being run on the OAM systems, or how often backups are run. Peak disk utilization can also be different from average disk utilization, for instance during backups or restore operations. While these guidelines are provided for the disk usage of the different VM types, customers should verify their disk usage under their own conditions since they are more driven by how the customer uses their system than by easier to calculate factors such as CPU utilization per MPS.

The OCDSR has been designed as a low disk utilization application, with all critical call processing applications performed in memory. There is also no swap disk utilization in any of the VMs. As a background for all of these numbers, the OCDSR has been run for years on "bare metal" deployments with a single pair of industry standard 10k RPM, 2½ inch disk drives in Raid 1. So even maximum sized OCDSR configurations run successfully on the approximately 120 IOPs provided by those disks. When run on higher performance disk subsystems such as SSDs, high disk utilization tasks such as background report generation just complete faster. The notes for each VM type give some of the factors that can drive different disk utilization levels. For instance, the primary traffic handling VMs, the IPFEs and the different types of MPs, have a fairly constant disk utilization independent of the traffic level. This is because the primary disk utilization is for saving statistics, then forwarding them to the SOAM.

Table B-1 VM Disk Utilization Characteristics

VM Name	Disk (GB)	Routine Disk Utilization (IOPs)1	Peak Disk Utilization (IOPs)2	Disk Usage Modes	Notes
DSR NOAM	120	100	800	Periodic (30 second) small writes to collect statistics. Large block reads to run reports and backups.	Background disk utilization is mostly statistics collection from managed DSRs. Peak disk utilization driven by customer report generation and maintenance activities such backups.
DSR SOAM	120	100	800
DAMP	120	50	500	Writes statistics to disk at 30 second intervals, reads them at 5-minute intervals to send to SOAM	Disk utilization is independent to traffic levels. Is affected by the size of the DSR configuration (number of connections for instance) and the utilization of features that create measurements such as ETGs and TTPs.
vSTP MP	120	50	500		Disk utilization is independent to traffic levels. Is affected by the size of the vSTP configuration such as the number of local and remote peers.
SS7 MP	120	50	500		Disk utilization is independent to traffic levels. Is affected by the size of the DSR configuration
IPFE	120	20	100		IPFE has relatively few configuration items and statistics, and very low disk utilization.
SBR(s)	120	50	800	Disk writes mostly short bursts for statistics storage	Peak disk utilization driven by recovery activities between active/standby servers.
SBR(b)	120	50	800
SDS NOAM	350	100	800	Synchronizes changes to in-memory database to disk. Mostly write application	SDS can maintain multiple copies of large subscriber database. Peak disk utilization is mostly driven by creating new backups.
SDS SOAM	175	50	800
SDS DP	175	80	500
SDS Query Server	350	100	800	Synchronizes changes to in-memory database to disk. Reads are driven by customer queries	The query server is not a real-time system. The amount of disk reads is driven entirely by manual customer database queries.

The "routine" disk utilization is the minimum engineered IOPs for the proper functioning of the VM. Average disk utilization is typically lower.

The "Peak" disk utilization is number of IOPs the VM is capable of using given sufficient resources.