Foglight: Turn key monitoring with automation feature set for environments with modest customization
Use Cases and Deployment Scope
- [Foglight is] used to monitor virtualized environment (VMware) Active Directory, Exchange, SQL Server w/ Analytics and other physical servers and software subsystems.
- The problems it solves for are active monitoring and alerting (uptime monitoring), historical baseline and trending, surfacing changes to hardware/software, log aggregation and analysis.
Pros
- Graphing and visualization of complex data sets
- Historical baselining and trending
- Automation of common tasks for response to known issues
- Aggregation and correlation of data across time series
Cons
- Threshold limits for alerting are not easily set to custom instances or groupings (things like customizing free disk space alerts for systems with enormous volumes to alert at a different threshold % than those with smaller volumes)
- Groovy requires a learning curve/specialized skill set for modifying rule sets
- Quest (perhaps as a victim of being purchased and then subsequently sold off by Dell) has deprecated a number of the subsystem cartridges (Storage etc.) that we depended upon including the lowest common denominator one (snmp) that is used to create monitors around proprietary systems and hardware that has no other api/monitoring capability
- Support has in past taken *months* to get an answer to issues that they had already solved for under another cartridge (e.g. disabling of SMB v1 caused their Exchange cartridge to have monitoring failures - they corrected this. Odd issues with another cartridge took literally 6 months to figure out that these were related to the disabling of SMB v1 as well. This was very frustrating.
- Other products offer more advanced analytics so it seems they have been falling behind. If this were a very inexpensive product I would not fault this but the licensing for Foglight is not inexpensive.
Likelihood to Recommend
- Monitoring, Alerting and Management of on-prem virtualization and physical servers
- Historical trending and baselining are excellent and being able to compare a problem period against the same sets of data for a period when there was no issue is especially helpful as part of troubleshooting.
- Ease of use for end users is appreciated
- Creation and management of custom monitors is a developer level activity requiring Groovy skillset
- Less useful for management of an environment where there is a large amount of custom implementations
- Commitment to longevity of cartridges/packs for specific subsystems
- Ability to create custom schedules/alerts/escalations is only available via a 3rd party product (from a former Quest developer) and is not particularly easy to manage/manipulate
- Lack of API accessibility for external integrations to other analytics (splunk, etc.)
