Heartbeats Corp. Publishes Server Monitoring Case Study for FuRyu's Large-Scale Entertainment Service 'Pictlink'

Heartbeats Corp. (Headquarters: Shinjuku-ku, Tokyo; CEO: Masanori Fujisaki), a provider of fully managed server monitoring, maintenance, and operation services, has published a case study detailing its collaboration with FuRyu Corp. (Headquarters: Shibuya-ku, Tokyo; President & CEO: Masahito Enomoto). The study focuses on the implementation of Heartbeats' 'Initial Server Monitoring Response Service' for 'Pictlink', FuRyu's service for retrieving and viewing images taken at photo sticker booths.

Through this project, initial response times during system outages—which previously could take up to several hours—have been significantly shortened. By reducing the frequency of incident handling for on-site engineers, the company has established an environment where teams can concentrate on developing new features and continuous improvement activities.

Background of Implementation: Challenges with Individual Dependency and Delayed Responses Post-Cloud Migration
FuRyu's 'Pictlink' is a beloved entertainment service primarily targeting teenagers and young adults. The SRE team responsible for the service's infrastructure faced major challenges after migrating from an on-premises setup to a multi-cloud architecture utilizing both AWS and OCI.

- Ambiguous Response Workflows: Although automated Slack notifications were in place, post-alert procedures were not well-defined, leaving it unclear who should respond when an outage occurred.
- Load Concentration on Specific Members: This led to delayed initial responses and a continuous situation where specific team members bore the brunt of incident handling.
- Weekend On-Call Burden: During peak seasons, developers were required to be on standby during weekends, heavily increasing the operational burden.

Deciding Factor for Choosing Heartbeats: Faster Situation Awareness and Initial Response via Human Phone Notifications
To resolve these issues, FuRyu adopted Heartbeats' 'Initial Server Monitoring Response Service'. The key factors in their selection included:

- Human-Operated Phone Calls: Unlike automated machine voices, real personnel assess the situation and provide verbal phone notifications. This makes it easier to grasp the context, leading to faster decision-making.
- 24/7/365 Stable Support System: The reliability of entrusting operations to professionals, thereby eliminating internal weekend on-call duties.

During the onboarding process, both companies collaborated to configure monitoring items and revise alert thresholds, which Heartbeats ultimately compiled into a 'Monitoring Specification Document'. FuRyu now utilizes this document internally to share current monitoring protocols.

Achievements Post-Implementation: Accelerated Initial Responses and Resource Reallocation to Core Development
- Significant Improvement in Initial Response Speed: Heartbeats aims for a Service Level Objective (SLO) of responding within 10 minutes. This structure drastically reduced the time from a 'Pictlink' outage to the initial response, fortifying the service's reliability.
- Reduction of Operational Load and Elimination of Weekend Standbys: Weekend on-call duties during busy periods are no longer necessary. With Heartbeats handling the initial response entirely, the frequency of interruptions for internal engineers has dropped.
- Focus on Core Tasks: Engineers can now allocate much more time to their primary responsibilities: developing new features and driving improvement initiatives.

Comments from FuRyu Corp. Project Managers
For the publication of this case study, representatives from FuRyu Corp. shared the following insights:

-- Could you share a specific episode where you felt glad to be using Heartbeats' services?
Mr. Kushima: 'Shortly after implementing Heartbeats' service, we requested an irregular change—asking them to alter contact destinations depending on the specific application. They carefully listened to these detailed requests and willingly accommodated us, for which we are deeply grateful.'

Mr. Awata: 'During the peak season last year, when we adjusted our auto-scaling settings, we requested an increase in monitoring thresholds. They responded immediately after we contacted them via Backlog, which was incredibly helpful.'

-- What do you prioritize most in the process of selecting an outsourcing partner?
Mr. Awata: 'The balance between cost and service quality. For that reason, to mitigate risk initially, we contracted Heartbeats for a limited 3-month period. After actually experiencing the strong benefits of human-operated responses, we decided to continue the partnership. Automated machine voices can sometimes make it difficult to understand the situation, but having a person directly explain it makes a huge difference.'