MetGlobal: three times the capacity at two-thirds of the cost with Google Cloud Platform
By developing technology products for the travel industry, MetGlobal has become a major player in B2B travel e-commerce. Housing a portfolio of major brands including HotelsPro, Otel.com, and tatil.com, MetGlobal handles a global inventory of over 500,000 hotels at 20,000 destinations worldwide. Prices and availability change constantly, so MetGlobal stays competitive with ultra-fast data processing on Google Cloud Platform.
“Our previous cloud provider went against the logic of cloud computing. By making us pay up front and predict our capacity in a changing environment, it created excessive costs and a heavy operational load that limited our ability to optimise resources. On top of that, it just didn’t have the speed to process data in real-time. Google Cloud Platform is more affordable, runs itself and gives us the tools we need to reach our goals.” - Gizem Terzi Türkoğlu, Project Coordinator, MetGlobal
More speed, less DevOps
The online hotel booking industry is highly competitive, so businesses aim to offer the fastest access to the largest inventory of hotels available. Every day, the HotelsPro system alone processes 100 million lines of incoming data and sends 250 million lines to its suppliers, while MetGlobal processes terabytes of data over more than 5,000 xml connections. MetGlobal looked to increase processing speed both to improve customer experience on its portfolio and partner websites and to enable real-time analytics for business intelligence.
To achieve those goals, MetGlobal worked with Bulut Yazılım to switch to Google Cloud Platform. Using CloudEndure block level replication tools, Bulut Yazılım helped migrate more than 600 VMs to Google Cloud Platform in less than 90 days. With Google Cloud Load Balancing and autoscaling on Google Kubernetes Engine, MetGlobal reports a response time of less than one second across a global network. Meanwhile the MetGlobal business intelligence team uses Google Cloud Dataflow,Google Cloud Dataproc and Google BigQuery to enable fast and powerful analytics with minimal DevOps. By switching to a serverless environment and microservices infrastructure and using GCP products for everything from sharing data with partners to personalising user data on websites, MetGlobal has sharply reduced its operational load, freeing engineers to spend more time adding value to the business. Crucially, the faster, more efficient infrastructure cuts costs without the need for long-term commitments, thanks to Google’s innovative pricing and sustained use discounts.
“With Bulut Yazılım and help from the Google sales team, we moved the majority of our platforms in four weeks. Bulut Yazılım then migrated the legacy PostgreSQL database to GCP manually, using CloudEndure’s continuous block-level replication tools to move an additional 32 servers from the legacy platform in just two weeks. In total, the entire migration including more complex legacy systems was complete within 90 days.” - Gizem Terzi Türkoğlu, Project Coordinator, MetGlobal
Saving 30% on costs, tripling capacity
Switching to GCP decreased MetGlobal infrastructure costs by 30%, while tripling data processing capacity so that the business intelligence team can analyse 2TB of data a day. MetGlobal now looks to use additional GCP tools to refine its offering, including Vision API to expand its usable inventory by detecting and deleting duplicate images on its vast database. In the long-term, the company plans to apply machine learning to handle its cache of room availability information and intelligently dispose of outdated data.
“Knowing that Google has invested a lot in AI, and knowing that it shares its knowledge with its partners, was a major factor in our decision to choose Google. We hope to leverage that huge investment with our BI department so that we can make the most of big data and the latest in data science.” - Gizem Terzi Türkoğlu, Project Coordinator, MetGlobal