Shown below is a diagram of sizing-estimation process. It is not the only method for sizing deployment, but it should provide useful insight and guidance
1. Estimate user concurrency and throughput
In most use-case scenarios, the number of expected concurrent users is far less than the number of named users of system. A scenario in which all of the named users are simultaneously accessing the system is extremely rare. Normally, a globally distributed organization with users spread out all over the world experience about ⅓ of its named users online at any time during their individual 8-hour work days.
Concurrency typically ranges from 10% to 50+% for App deployments.
10 % [ Minimum ]
500000
50000
20 % [ Average ]
500000
100000
30 % [ Recommended ]
500000
150000
40 % [ Highly Scaled ]
500000
250000
2. Estimate SeMA Application HW sizing
A- Every CPU core handles 2 Worker
B- Worker can handle 120 Portal users
C- Concurrent user = Users use app simultaneously with time session 5 sec
D - Best practice of workers per machine from 17 to 33 [Due to processor queuing]
Number of portal Users [500000]
Number of concurrent users = 40 * 500000 / 100 = 200000
× Number of Woker = 200000 / 20 = 100 Woker
x Number of Workers / VM = 33
x Number of Core / VM = 16 core
x RAM cacpity estimation [ 1 core * 4 ] = 16 * 4 = 64 Gb / VM
Minimum 4 VM with 16 core / 64 Gb RAM
Recommended 5 VM x with 16 core / 64 Gb RAM
3. Estimate SeMA DB HW sizing
Database storage requirements may be estimated roughly using the following calculation:
Number of Users [ 5000 ]
× Number of events per host per day [ 15 ]
× 5Kb per event
Number of Users × Number of events per host per day
× 5Kb per event
For example, an organization of 5,000 Users, with each user generating
an average of 15 events per day,
requiring a 30 day retention would require a database capacity of:
## DB stoage sizing / Month
5,000 × 15 × 5 × 30 = 11,250,000Kb, or 11Gb
## DB stoage sizing / Year
11 GB per Month x 12 = 132 Gb / Year
DB HW sizing for 5000 client connections / sec
# DB Version: 14
# OS Type: linux
# DB Type: web
# Total Memory (RAM): 96 GB
# CPUs num: 24
# Connections num: 5000
# Data Storage: san
max_connections = 5000
shared_buffers = 24GB
effective_cache_size = 72GB
maintenance_work_mem = 2GB
checkpoint_completion_target = 0.9
wal_buffers = 16MB
default_statistics_target = 100
random_page_cost = 1.1
effective_io_concurrency = 300
work_mem = 1258kB
min_wal_size = 1GB
max_wal_size = 4GB
max_worker_processes = 24
max_parallel_workers_per_gather = 4
max_parallel_workers = 24
max_parallel_maintenance_workers = 4
4. Estimate SeMA File Storage HW sizing
x Number of users [ 5000 ]
× Number of survey per user per day [ 2 ]
× 25 Mb per survey attachment [ 25 mb ]
Number of Users × Number of survey per user per day
× 25 Mb per survey attachment
## File storage sizing / Month
5,000 × 2 x 25 Mb × 30 = 750 Gb
## File storage sizing / Year
750 GB per Month x 12 = 9 TB / Year
SeMA App file storage can start with 4 TB and extend based on monitoring actual usage of storage
