Nearly two weeks have passed since we released the latest version of EasyDCIM. How time flies! V1.3.0 is featured with tons of amazing changes, among which full support for PHP 7 is definitely one of the most interesting ones. Heading towards improving Admin Experience in our system, we have begun with a tremendous performance boost. Not to leave that seemingly bold statement in limbo, we have prepared a detailed comparison of various loading times between the current version and the previous one. Grab a cup of hot tea, coffee or anything you like and take a look at hard facts on how much time you will save with new EasyDCIM v1.3.0.

Our tests have been carried out on an identical machine with the same measuring methods: DOMContentLoaded and load by MDN.

OVH Machine:

  • KVM OpenStack
  • 1 vCore
  • 2,4 GHz
  • 2 GB RAM
  • SSD 10 GB
  • Local Raid 10

Testing Environments:

  1. Ubuntu 14.04.5 LTS + PHP 5.5.9 | 100 servers + 10 switches (all with SNMP support)
  2. Ubuntu 16.04.2 LTS + PHP 7.0.15 | 100 servers + 10 switches (all with SNMP support)

 

Table for DOMContentLoadedTable for load

Seconds Per Page
Ubuntu 14.04.5 LTS + PHP 5.5.9 Ubuntu 16.04.2 LTS + PHP 7.0.15 Ubuntu 14.04.5 LTS + PHP 5.5.9 Ubuntu 16.04.2 LTS + PHP 7.0.15
Servers List
(10 per page)
5.52 2.08 6.51 3.20
Servers List
(20 per page)
6.51 3.54 10.56 5.42
Servers List
(50 per page)
14.33 4.12 33.94 9.25
Server Summary Page 2.29 2.18 3.13 2.57
Server Network Interfaces 2.35 1.90 2.90 1.94
Network Devices List (10 per page) 2.57 1.96 6.88 3.34
Dashboard 3.67 2.20 4.12 2.23

Now let’s take a glance at the charts below to visualize the magnitude of changes.

DomContentLoaded - EasyDCIM

Load - EasyDCIMIt is apparent that new EasyDCIM v1.3.0, which fully utilizes the potential of PHP 7, works much faster than its predecessor – v1.2.3. As you can see, in several instances the loading time of pages is even more than two times shorter! What will you do with that extra time saved each day? From now on it is totally up to you!

Author
Comments
Write a reply