✍ Kashif M ™
☎ +91 9994883085
✉ kashif7222@gmail.com


Tuesday, 3 April 2018

Elastic search - Kibana Architecture

By on April 03, 2018

KIBANA Architecture

Difference between BI and KIBANA:

                       BI                                                                                                 KIBANA
1)      BI will connect directly to the database through ODBC or OCI ….etc
1) KIBANA will use ElasticSearch and this           Elasticsearch loads data from database.
2)      BI can load different types of databases or schemas In to the RPD
2) ElasticSearch cannot load different types of databases or schemas at a time
3)      After importing schemas of different databases, we will define joins for Fact and
Dimensions in the RPD according to our requirement.
3) In Elasticsearch we have to give query with well defined joins for entire schema.
4)      We can do complex quries  or calculations at RPD level and report level
4) We can’t do any calculations in report level except total, min, max, mean.
5)      We can develop dashboards with different reports.
5) We can develop dashboards with Histogram, Pie and Table formats only.
6)      We can develop reports with different group by clauses.
6) We can’t develop reports with different group by clauses.
7)      For Graphs we can change x-axis and y-axis
According to our requirement, both x-axis and y-axis can be numeric or character field.
7) But in KIBANA we can’t change x-axis, by default it is TimeField, y-axis --- it will accept only numericfields.
8)      We can create graph with ‘n’ number of columns.
8) We can’t…..
9)      In BI Pie chart will generate based upon measure fields and character fields.
9) In KIBANA Pie chart will generate with only one field and it will take count of that field.
10)   Granularity is possible in BI
10) There is no Granularity in KIBANA.
11)   We can navigate to one report to another report.
11) We can’t…..
12)   We can download reports in to PDF, CSV, Excel…..etc
12) We can’t……
13)   We can do Time series calculations for YearAgo, MonthAgo, and WeekAgo…
13) We can’t…..
14)   We can send reports

15)   We can create n-number of users for each roles
15)We can’t create users in KIBANA
16)   Data level and Object level security is possible…..
16)There is no security concept in KIBANA
17)   Business Intelligence content can be delivered to a variety of devices like web browser ,E-mail, Digital cell phone…etc
17) We can’t….

Major Drawbacks Of KIBANA:

1)      We can’t change graph properties of Kibana, for example for histogram by default it is taking x-axis as TimeField and y-axis as numeric fields.
2)      We can’t use multiple group by clauses.
3)      There is no Granularity in KIBANA
4)      Pie chart is taking only one field and displays count of that field.
5)      We can’t do any new calculations in report level.
6)      Query mode is taking for entire dashboard.
7)      We can’t do time series calculations.
8)      There is no security concept in KIBANA.

Below are KIBANA reports which are developed for sample… (I took only 3 months of Data)

Total Sales report:

In above report Pie chart is taking count of Month_ID field.

Total Sales Restaurant Group Wise:

Total Sales Restaurant Name Wise:

Total Sales Product Wise:

mkashu: Elastic search - Kibana Architecture
Review : Kashif | Kashif
Update: April 03, 2018 | Rating: 4.5

Comment for "Elastic search - Kibana Architecture"


Post a Comment