# Graph application response time latency in milliseconds

**URL:** <https://discuss.elastic.co/t/graph-application-response-time-latency-in-milliseconds/71792>\
**Category:** Kibana\
**Created:** [January 16, 2017, 11:18pm UTC](https://discuss.elastic.co/t/graph-application-response-time-latency-in-milliseconds/71792 "2017-01-16T23:18:58Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![mussa572](https://avatars.discourse-cdn.com/v4/letter/m/a698b9/32.png) [@mussa572](https://discuss.elastic.co/u/mussa572)\
**Post date:** [January 16, 2017, 11:18pm UTC](https://discuss.elastic.co/t/graph-application-response-time-latency-in-milliseconds/71792/1 "2017-01-16T23:18:59Z")

</div>

We have financial application which works on FIX protocol . users Sends order and our application sends the response back within few milliseconds , application also generate the log with processing time of each order .. Below is how we store Data in ES . We are interested generating the vertical graph for each FIX users with their corresponding OrderId and the time it took for our application to response back the order.

I want to understand what is the best way to convert and the milliseconds value to seconds or minute and show on the Y-axis with the time of order receive on the X-Axis .

Order Receive time , App response time milliseonds(us) , Order ID , Source IP , FIX USER

00:51:34.575 , 3860 , 36A301295, 192.168.1.1 , ABCFIX1  
00:52:34.575 , 41256 , A12342524, 10.168.1.1 , XYZFIX2

---

<div class="post-metadata">

**Author:** ![spalger](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/spalger/32/14092_2.png) [@spalger](https://discuss.elastic.co/u/spalger)\
**Post date:** [January 16, 2017, 11:42pm UTC](https://discuss.elastic.co/t/graph-application-response-time-latency-in-milliseconds/71792/2 "2017-01-16T23:42:35Z")

</div>

Kibana visualizations are intended more for viewing aggregate data, rather than the individual data points that make up the aggregate. Because of this you will need to use an `average` of the response time (which will match the single response time if you filter down your data enough).

I would probably start with a vertical bar chart, then:

1. "Split charts" by the `Terms` of the `FIX USER` field
2. "X-Axis" using a `Date Histogram` of the `Order Received Time` field
3. `Average` metric of the `App response time` field

When you are zoomed in close enough you will see individual orders on the chart, but as you zoom out and look at larger amounts of time the date buckets will start to cover multiple orders and the average response time will be based on all of the orders that fit into those buckets. Because of this you will be able to scale from a 10 minute time window to a year long time window

PS: you will probably want to expand your `Order Receive Time` to a more complete timestamp (including the date)

---

<div class="post-metadata">

**Author:** ![mussa572](https://avatars.discourse-cdn.com/v4/letter/m/a698b9/32.png) [@mussa572](https://discuss.elastic.co/u/mussa572)\
**Post date:** [January 17, 2017, 10:42pm UTC](https://discuss.elastic.co/t/graph-application-response-time-latency-in-milliseconds/71792/3 "2017-01-17T22:42:23Z")

</div>

Thanks for the reply Spencer  
what is the best way to Convert App response time from 3860 to like 3.8ms as this will make more sens for user to check the average based this format . Also is there any way where i can add the drop down menu where users can select the FIX user, which can then show the graph based on per user inthe same format which you mentioned on the last reply.

---

<div class="post-metadata">

**Author:** ![LeeDr](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/leedr/32/9289_2.png) [@LeeDr](https://discuss.elastic.co/u/LeeDr)\
**Post date:** [January 18, 2017, 11:07pm UTC](https://discuss.elastic.co/t/graph-application-response-time-latency-in-milliseconds/71792/4 "2017-01-18T23:07:07Z")

</div>

I can help with the milliseconds to seconds formatting. If you go to your index pattern in Kibana, and select the edit pencil icon for the field, you can set the  
`Format: Duration`  
`Input Format: Milliseconds`  
`Output Format: Seconds`  
click `Update Field`

 ![](https://us1.discourse-cdn.com/elastic/original/2X/4/4171484087a8ef8afe5adc61d1a49e99eed68774.png)

You can't add a drop-down menu, but if you add your visualization to a dashboard, and you also create a data table visualization of your "FIX users" and add that to the dashboard, you can sort of use it like a drop-down. When click click on one of the names it will add a filter that affects the whole dashboard. Then you can either delete or disable the filter. So you could disable, add another filter, etc, until you had the filters you wanted (unless there's a lot), and then enable which ever one you want. I realize its a clumsy work-around, but if you really need it you could try that.

This data doesn't demonstrate the technique very well, but you get the idea.

 ![](https://us1.discourse-cdn.com/elastic/original/2X/c/c2ecd4d4fe8a519e3d92fefeee0be717e7c070fe.gif)

Lee

---

<div class="post-metadata">

**Author:** ![system](https://us1.discourse-cdn.com/elastic/original/3X/1/a/1ac57faf039f6b580b3f104ef42a2a89e41014de.png) [@system](https://discuss.elastic.co/u/system)\
**Post date:** [February 15, 2017, 11:08pm UTC](https://discuss.elastic.co/t/graph-application-response-time-latency-in-milliseconds/71792/5 "2017-02-15T23:08:08Z")

</div>

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.
