IBM Tivoli Monitoring for Virtual Servers collects a wealth of
performance metrics for a virtual environment in the Tivoli Data
Warehouse when historical collection is enabled on the agent. Data can be collected over long periods of time at hourly, daily, weekly, monthly, quarterly and yearly intervals. When mined, this data helps analyze the overall performance and capacity of the environment in the past, present and future.
Using our very popular Tivoli Common Reporting
tool (which runs Cognos 8.4 Business Intelligence and Reporting suite underneath), we leverage the data in the Tivoli Data Warehouse to provide intelligent data models and reports that will help you understand your environment better and make basic estimates for capacity. The reports are not geared toward long-term capacity
planning, but are more a quick analysis tool for understanding work-load
characteristics, placement and optimization.
These reports answer various questions like
- How many more workloads can I add to this cluster or host based on historical usage?
- If I want to add X more workloads how much more resources will I need?
- Will I run out of resources like CPU, memory , storage within the next few days/weeks/months?
- Is my environment balanced?
are the top/bottom resource consumers in the environment? For e.g.,
- Can you
show me the top 10 VMs in this cluster which are consuming more than 90%
- Or can you show me the bottom 5 hosts in a cluster by CPU ,
memory, storage, network or disk I/O?
- Or can you show me the top Clusters in my datacenter by available CPU/memory/storage?
- What are the top physical NICs consuming bandwidth?
- Which host servers are overcommitted for memory?
- Which datastores in the past week have had maximum available space and low disk I/O and what VMs were connected to it?
- Are my hosts overloaded by CPU leading to large wait times for the VMs? I want to be able to compare CPU utilization of VM and host, percent ready of VM and number of VMs on the host in one chart
- Show me the VMs that peak together by comparing their key metrics over time
- Did the utilization of key resources or configuration of the environment change by X% between one week and another?
- Show me monthly utilization of key resources for Cluster/Server/VMs and let me drill down from month to week to day to hour.
The TCR tool is completely web based and has a bunch of interesting features that will simplify your life:
- It allows you to view and administer reports in
various formats like HTML, PDF, Excel, CSV etc.
- It lets you create reports by drag and drop without having to write any SQL queries.
- The tool is web based , hence you can edit a report in the browser while viewing it and the changes will be saved right then
- Reports can be emailed and also conditionally emailed based on thresholds
- Reports can scheduled to run at regular intervals
- Reports are interactive and allows users to drill through from one
level to another. For example, it allows drill downs in context from cluster level
data to host server level to VM level. It also allows drill downs from
one frequency to a more granular frequency in time like from Monthly to
Weekly to Daily to Hourly data.
These type of reports are available for the VMware VI agent (shipped with ITM for Virtual Servers v6.2.3), System P agent (on ISM Library), Citrix XenServer agent (on ISM Library) etc.
To learn more take a look at our video demos
and white paper
Payal Chakravarty, Technical Lead, Reporting Solutions, Tivoli Software