As the lead technical strategist at a specialty industrial controls provider that’s spent the last decade delivering HJS OPS-C solutions to energy and manufacturing clients across three continents, I get asked a lot of one-sided questions about this platform: Is it easy to install? Does it reduce downtime? But over the years, the question that’s tripped up more of our clients than any other, especially early on, is: What are the actual data storage requirements for HJS OPS-C? Most teams approach this thinking it’s just another monitoring tool with generic storage needs—until they get hit with unplanned downtime, overprovisioned cloud bills, or even incomplete operational audits because their storage didn’t line up with how HJS OPS-C actually works. Today, I’m pulling back the curtain on the storage rules we’ve developed from supporting 120+ active HJS OPS-C deployments, to help you avoid the costly missteps many of our first-time customers made. HJS OPS-C
First, let’s ground this in what HJS OPS-C actually does, because every storage decision ties back to its core functions. This is not a basic SCADA tool for logging daily plant temperatures. It’s a real-time operations optimization suite built for heavy industrial environments—think combined-cycle power plants, steel rolling mills, and chemical processing facilities. It ingests time-series data from hundreds or thousands of connected sensors every second: pressure transducers, vibration monitors, flow meters, and control system modules. That data isn’t just stored; it’s processed on the fly to run predictive analytics, detect anomalies (like a pump’s vibration spiking before it fails), generate compliance reports for regulatory bodies (OSHA, EPA, or ISO), and sync with existing plant ERP systems to align operations with supply chain and maintenance schedules. For context, a mid-sized power plant running HJS OPS-C will generate ~2.2 terabytes of raw time-series data per month, just from its core sensor network. That number jumps to 5 TB for large-scale facilities with redundant sensor arrays. That’s not the kind of data you can tuck into a shared server drive and forget about, which is why storage for HJS OPS-C has three distinct phases, each with non-negotiable requirements: hot storage, warm storage, and cold storage.
Let’s start with hot storage—this is the active storage pool that HJS OPS-C accesses every single day, for real-time operations, current analytics, and immediate troubleshooting. The rules here are non-negotiable, and they’re different from what most industrial IT teams are used to for other tools. First, hot storage needs to be low-latency, with a maximum read/write latency of 10 milliseconds or less. HJS OPS-C’s core real-time modules—like its anomaly detection engine and operator dashboard—depend on pulling data from the last 72 hours to flag live issues. If latency creeps above 10ms, the tool can misclassify a critical vibration spike as noise, or delay alerting to the control room by enough time to trigger unplanned downtime. We saw this first-hand with a 2022 client in the U.S. Midwest: their team used a standard HDD array for hot storage to cut costs, and experienced 35ms latency during peak production hours. The result? HJS OPS-C missed a pump failure that caused $1.2 million in unplanned maintenance. Beyond latency, hot storage retention is strict: HJS OPS-C requires a minimum of 7 days of uncompressed raw sensor data stored here, plus processed analytics data for the same window. For most deployments, this works out to a minimum hot storage allocation of 1 TB for small facilities (100–500 connected sensors), 4 TB for mid-sized facilities (500–2,000 sensors), and 8 TB for large facilities (2,000+ sensors). Wait—you might be wondering, why not compressed data here? Because even lossless compression adds a small processing overhead, and for hot storage, every millisecond of extra processing counts. We advise clients to use enterprise-grade SSDs for hot storage, configured in a RAID 10 array for redundancy—no exceptions. A RAID 5 array works for non-critical files, but for HJS OPS-C hot storage, a single disk failure could corrupt real-time data mid-operation, and the cost of downtime far outweighs the slightly higher cost of RAID 10.
Next is warm storage, the middle tier of storage for data that’s not needed in real time, but needs to be accessed regularly for weekly performance reviews, monthly maintenance planning, and internal operational audits. This is where most clients cut corners that come back to haunt them. Warm storage for HJS OPS-C has two key requirements: it needs to support efficient compression (since most of this data is no longer needed at full detail) and it needs to retain data for a minimum of 12 months, per HJS OPS-C’s architecture specifications. The compression here can be lossless, so you don’t risk degrading historical sensor data, and HJS OPS-C’s built-in tools are optimized to parse compressed time-series data without adding latency during access. For example, if your plant needs to compare this month’s pump performance to the same month last year, HJS OPS-C pulls the compressed warm storage data, decompresses only the relevant time windows, and runs the analysis without affecting hot storage performance. We recommend nearline SAS HDDs or even NVMe SSDs for some clients with high query volumes here, because they offer a better balance of speed and cost than high-performance SSDs. The key number here: warm storage allocation should be at least 12x your monthly raw data throughput, plus 20% overhead for processing and metadata. For the mid-sized power plant I mentioned earlier, that meant ~26.4 TB of warm storage (2.2 TB x 12) for their first year, which was a fraction of the cost of overprovisioning hot storage to hold a full year of data. A common mistake here is trying to store warm data on cloud object storage without configuring the right access policies. We’ve had several clients test AWS S3 for warm storage only to find that they had to increase their query timeout to 30 seconds, which broke HJS OPS-C’s automated report generation for weekly performance reviews. If you’re using cloud for warm storage, make sure you use a tiered storage service with low-latency retrieval for frequently accessed objects, not standard cold cloud storage.
Then there’s cold storage, the archive tier for data that’s only needed for long-term compliance audits, regulatory reporting, or post-incident investigations. This is the cheapest tier, but it still has specific requirements for HJS OPS-C. First, cold storage needs to retain data for a minimum of 7 years—most global regulatory bodies (like the EPA in the U.S., EUNOMIA in the EU, and ISO 50001 for energy management) require industrial facilities to retain operational data for 5–10 years for audit purposes. HJS OPS-C’s native data export tool can pull full 7-year datasets in a standardized, non-proprietary format (CSV or Parquet) that’s easy to share with auditors, so you don’t want to have to reformat or recover this data later. The other key requirement here is that cold storage needs to be write-once, read-many (WORM) compliant. That means once data is written, it can’t be modified or deleted for the full retention period—this prevents bad actors (or accidental human error) from altering operational data to falsify compliance reports. We’ve seen clients get fined $200,000+ in EPA audits because they tried to delete or modify 3-year-old sensor data to fix a minor discrepancy, not realizing that HJS OPS-C’s audit trail would flag the change. For cold storage, we recommend either on-premise WORM-compliant tape libraries for clients that don’t want to rely on third-party cloud providers, or cloud WORM storage tiers like AWS S3 Glacier Flexible Retrieval, configured with 7-year lock policies. The allocation here is straightforward: cold storage should be 84x your monthly raw data throughput (7 years x 12 months). For that mid-sized power plant, that’s ~184.8 TB of cold storage over 7 years—again, a fraction of the cost of storing all data on high-performance SSDs long-term.
Now, let’s talk about the common edge cases that most storage guides for HJS OPS-C leave out—things that trip up even teams that get the basic three-tier model right. First, time zone alignment and data timestamping. HJS OPS-C relies on precise, synchronized timestamps from sensor NTP servers to correlate data across different systems. If your storage system is configured in a different time zone, or if timestamps are off by even a second, you’ll end up with gaps in your analytics or incorrect audit reports. We once had a steel mill client in India that set their cold storage timezone to UTC, while their sensors were set to Indian Standard Time (IST). Over the course of a year, that created a 5.5-hour gap in their data alignment, leading to a failed ISO 50001 audit that cost them a major energy efficiency certification. The fix was simple: sync all storage tiers to the same timezone as your control system’s NTP server, and use HJS OPS-C’s built-in timestamp validation tool to cross-check every quarter.
Another edge case: redundant storage for geographically dispersed HJS OPS-C deployments. If you’re running HJS OPS-C across multiple plant sites, you’ll need a centralized storage layer that syncs data from all sites, with the same tiered structure. For example, a large energy provider with 8 power plants might have each site run its own hot and warm storage, while all cold storage is archived to a central WORM system. The key here is to set up data replication for hot and warm storage to a secondary site, to avoid data loss in case of a site power outage or server failure. We usually recommend synchronous replication for hot storage (so every write to the hot tier is copied to a secondary site in real time) and asynchronous replication for warm storage, which is fast enough without adding latency for real-time operations.
And let’s not forget about scalability. HJS OPS-C is designed to grow with your facility—if you add 500 new sensors to your plant next year, the storage requirements will scale proportionally. Most of our clients make the mistake of overprovisioning storage upfront, which wastes capital, or underprovisioning, which leads to emergency upgrades mid-production. Our standard recommendation for clients is to use modular storage systems that let you add SSDs, HDDs, or cloud storage tiers as needed, rather than a fixed, one-time storage purchase. For example, if you’re a small plant with 200 sensors, you can start with 1 TB of hot storage, 12 TB of warm storage, and ~180 TB of cold storage—then add another 100 GB of hot storage, 6 TB of warm storage, and 90 TB of cold storage when you add 500 new sensors. This modular approach ensures you only pay for the storage you need, when you need it.
Over the past decade, we’ve learned that the biggest mistake clients make with HJS OPS-C storage is treating it as an afterthought, rather than a core component of the platform’s functionality. A great HJS OPS-C system is only as good as the storage that supports it—if your storage is too slow, too small, or not compliant, you’re not getting the full value from the platform’s predictive analytics, audit tools, or optimization features. We’ve seen clients go from $1.2 million in downtime to $3 million in annual cost savings just by optimizing their storage tiers for HJS OPS-C, without changing anything else about their operations.
If you’re in the process of planning a new HJS OPS-C deployment, or you’re looking to upgrade your existing storage to support your current platform, our team of technical experts can walk you through a customized storage assessment, tailored to your facility size, industry, and regulatory requirements. We don’t sell one-size-fits-all storage solutions—we work with you to map your specific sensor count, data throughput, retention needs, and compliance rules to the three-tier storage model that works for your operations. Whether you need on-premise storage, cloud storage, or a hybrid solution, we can help you design a system that keeps your HJS OPS-C running reliably, meets all regulatory requirements, and avoids the costly missteps we’ve seen from our clients over the years. Reach out to our team to start a conversation about your HJS OPS-C storage needs, and let’s make sure your next deployment is a successful one.
HJS OPS-C References:
International Society of Automation. (2021). Time-Series Data Management for Industrial Control Systems. ISA Press.
HJS Group Technical Documentation. (2023). HJS OPS-C Platform Storage Requirements v2.1.
U.S. Environmental Protection Agency. (2022). Industrial Operational Data Retention and Compliance Guidelines. EPA Office of Compliance.
ISO 50001:2018. Energy Management Systems – Requirements with Guidance for Use. International Organization for Standardization.
Shenzhen HJS Technology Group
Shenzhen HJS Technology Group is one of the most experienced hjs ops-c manufacturers and suppliers in China, featured by quality products and good service. Please rest assured to buy CE approved hjs ops-c made in China here from our factory. For price consultation, contact us.
Address: F10, Wanhe Technology Building C, No. 7 HuiTong Road, Fenghuang Street, Guangming New District, Shenzhen. China.
E-mail: info@hjs-pc.com
WebSite: https://www.hjs-pc.com/
