Finding bottleneck on my server

Install / Update Deployment
1

Dear All
feeling my server is sluggish with below specification on a dedicated server which seems to be slower than digital ocean’s USD20 package.

CPU : Intel(R) Atom™ CPU C2750 @ 2.40GHz
Ram 8GB
HDD: 160GB SSD
Core: 8

Checked the %iowait (IO) and %idle (CPU) which seems to be healthy. Also tuned mariaDB as below:

innodb_buffer_pool_size = 6G
innodb_buffer_pool_instances = 6
innodb_log_buffer_size = 8M
innodb_file_per_table = 1
innodb_open_files = 400
innodb_io_capacity = 400
innodb_flush_method = O_DIRECT
innodb_thread_concurrency = 0
innodb_read_io_threads = 8
innodb_write_io_threads = 8

are there something else i should be looking at? Thank you.

2

Are you sure it’s the server and not the bandwidth of the server? The specs look alright.

3

how to i measure that bandwidth info?

4

Try to download a big file off the server. Also ping the server. See if you’re dropping packets / ping is high.

1 Like
5

PING (62.210.81.50): 56 data bytes
64 bytes from 62.210.81.50: icmp_seq=0 ttl=57 time=131.092 ms
64 bytes from 62.210.81.50: icmp_seq=1 ttl=57 time=130.106 ms
64 bytes from 62.210.81.50: icmp_seq=2 ttl=57 time=130.281 ms
64 bytes from 62.210.81.50: icmp_seq=3 ttl=57 time=130.430 ms
64 bytes from 62.210.81.50: icmp_seq=4 ttl=57 time=129.556 ms
64 bytes from 62.210.81.50: icmp_seq=5 ttl=57 time=129.614 ms
64 bytes from 62.210.81.50: icmp_seq=6 ttl=57 time=129.131 ms
64 bytes from 62.210.81.50: icmp_seq=7 ttl=57 time=129.456 ms

bad?

6

Nah, it’s file. Did you check bandwidth?

7

downloading a 5G file off server, averaging 500KB/s , think thats fine?

8

Use htop - an interactive process viewer

9

10

@Jay try it when you are experiencing slugglish behaviour

That is a very underused server at the moment. Maybe you have network latency (?)

11

@Jay Yea, that’s not bad

@rmehta The latency is ~130, which isn’t terrible

12

ok will try it when server on heavy load,thank you for your time