Speaking at the 2011 PNEC conference in Houston this month Randy Petit reported successful trials of two open source tools in Shell. First up was Nagios (www.oilit.com/links/1105_10), a cross platform IT infrastructure monitoring tool that tracks multi-protocol network traffic and provides configurable alerts on failures, bottlenecks and other issues allowing for timely remediation before outages affect the business.
Shell uses Nagios for ‘proactive’ monitoring of connections to its external Petrobank data server hosted by Halliburton’s Landmark unit. Nagios is used to report issues and plan downtime. Nagios also provides ‘critical infrastructure’ monitoring of Shell’s own key systems, applications and databases. Shell needs to know more than just ‘the Apache server is up.’
Shell’s geomatics department provides a Nagios-powered RSS feed to keep users posted on network activity. Standard action plans have been developed providing who to inform and what actions to take in the event of an outage. All data issues can be reported to a single email alias. A dashboard shows system status with alerts sent out to pagers or Blackberries.
Nagios has extensions for license management, disk space monitoring and Oracle usage. The RSS Feed is used to alert staff to ‘on demand’ events like data loads, Recall refreshes and data orders. RSS can be consumed by ‘just about any device’ and provides an entry point to Shell’s SharePoint portal.
The Nagios/RSS combo has freed up Shell’s data managers for ‘proper’ work while the single mailbox has reduced disruption and duplication of effort. Nagios is considered to be the ‘industry standard’ for monitoring network services.
Petit’s next presentation covered the use of JTrak (www.oilit.com/links/1105_11) in Shell. JTrac, an open source development by one Peter Thomas, is described as a web-based ‘issue (i.e. bug) tracking’ application. But Petit spotted the opportunity to extend its use to upstream workflow orchestration. Target workflows include well data tracking, seismic requests and compliance with company procedures in these fields. Shell’s previous workflow management attempts were spreadsheet based and proved somewhat unsustainable.
JTrac plugs into existing systems and can perform Oracle or mySQL queries and inserts. JTrack provides notification to stakeholders at key junctures in the workflow with emails, and updates metadata as appropriate. JTrac manages roles and ‘who does what,’ passing information along to the ‘next allowed state,’ forcing completion and approvals. ‘States’ can be defined as mandatory or optional. The system generates a dashboard with ‘spaces’ for component workflows. Configurable KPIs provide management-level workflow tracking. Custom Java code allows interaction with the Landmark PowerHub API, linkage to the document management system and other data stores. Petit sees the JTrac workflow manager as a real ‘spreadsheet killer.’ Java makes the system multi platform allowing automated KPIs and metrics across Linux and Windows systems.
Comment—Nagios has good credentials amongst horizontal ‘blue chip’ users and oil and gas deployments including Landmark, Grant Prideco, Kelman, LMKR and Siemens. JTrac appears to be a more modest project. Perhaps this novel use in workflow orchestration will change that.
© Oil IT Journal - all rights reserved.