COMPUTE / HIGH MEMORY
Consider High Memory for database working sets, application caches and memory-intensive processing. A useful capacity estimate accounts for active data, concurrent processes and growth, rather than treating the total size of stored data as the memory requirement.
01 / Overview
Frequently accessed data and concurrent operations shape actual memory demand. A working set that does not fit may increase storage access, while unchecked connection or cache settings can consume available resources quickly. Read memory measurements alongside application behaviour before choosing High Memory.
Additional RAM does not by itself repair inefficient queries, memory leaks or connection management. Distinguish operating-system caching from application allocations and review per-process demand during busy periods. Aim for a clear operating plan as well as adequate capacity.
Size frequently accessed data, caches and active processes together. Do not assume that the size of every stored file must be matched with RAM.
Review connections, per-query memory and temporary operations. Test how individual settings affect aggregate consumption under realistic concurrency.
Separate memory demand from CPU pressure and storage waits. Use resource changes and software improvements where each addresses the actual constraint.
02 / Architecture
Illustrative working set: the application processes active data in memory and retains persistent data in storage. Memory is not a substitute for backup.
Selection guide03 / Selection guide
| Consider | How to decide | Why it matters |
|---|---|---|
| Active working set | Measure the data actually accessed within the relevant operating period. | Not every stored record needs to be resident in memory. |
| Concurrency | Combine connection and worker counts with per-process memory consumption. | A small allocation can become significant across many simultaneous processes. |
| Memory pressure | Monitor swap activity, reclaim behaviour and terminated processes. | A used-memory percentage alone does not explain the cause of a slowdown. |
| Growth and recovery | Plan for data growth and the warm-up period after a restart. | Capacity during recovery matters alongside steady-state operation. |
04 / Use cases
Review frequently read tables and indexes alongside concurrent queries. Validate sizing with representative data and the queries the application actually runs.
Define data lifetime and eviction behaviour. Plan how the application returns to the source when the cache is empty or unavailable.
Account for temporary copies created during transformation and analysis. Peak consumption during execution can matter more than the size of input files.
Plans & pricing
| Host | CPU | RAM | Storage | Bandwidth | Monthly price (USD) |
|---|---|---|---|---|---|
 FibaCloud
|
2
|
24 GB
|
20 GB
|
Includes 6TB bandwidth
|
28.00$Monthly
|
| Host | CPU | RAM | Storage | Bandwidth | Monthly price (USD) |
|---|---|---|---|---|---|
|
FibaCloud
|
4
|
48 GB
|
40 GB
|
Includes 8TB bandwidth
|
49.00$Monthly
|
| Host | CPU | RAM | Storage | Bandwidth | Monthly price (USD) |
|---|---|---|---|---|---|
|
FibaCloud
|
6
|
90 GB
|
90 GB
|
Includes 10TB bandwidth
|
74.00$Monthly
|
| Host | CPU | RAM | Storage | Bandwidth | Monthly price (USD) |
|---|---|---|---|---|---|
|
FibaCloud
|
10
|
150 GB
|
200 GB
|
Includes 12TB bandwidth
|
120.00$Monthly
|
Review the current configuration, price and order terms before selecting a service.
Choose a location to measure your connection latency.
05 / From planning to operation
Record application, database and operating-system consumption separately during normal and busy periods.
Inspect caches, connection pools and worker limits. Change one setting at a time and measure it under representative load.
Compare a High Memory plan using the same data and queries. Observe the effect on CPU consumption and storage activity.
Test restore and restart procedures. Assess how the application operates while caches are cold and data is loaded again.
No. The benefit depends on the working set and query behaviour. Locking, inefficient queries or write requirements may need other changes before memory capacity makes a difference.
Not necessarily. File caching can make productive use of available RAM. Review available memory, pressure indicators and application behaviour together.
It describes a resource family. Database installation, configuration, security, updates and backup responsibilities must be established separately.
RAM supports running processes and active data; storage retains persistent data. Measure data volume, the working set and storage access patterns as separate requirements.
Connect product decisions with installation, configuration and operating knowledge.
These references explain the technical concepts discussed on this page. Available configurations and service scope are defined by FibaCloud product and order information.
FIBACLOUD / High Memory
Tell us about your application, expected usage and operating requirements. Let’s bring the resource choices together.
