# Perspective Alarm Journal causes 100% CPU

**URL:** <https://forum.inductiveautomation.com/t/perspective-alarm-journal-causes-100-cpu/37804>\
**Category:** Ignition\
**Created:** [August 17, 2020, 1:02pm UTC](https://forum.inductiveautomation.com/t/perspective-alarm-journal-causes-100-cpu/37804 "2020-08-17T13:02:43Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![martin.johansson](https://avatars.discourse-cdn.com/v4/letter/m/7feea3/32.png) [@martin.johansson](https://forum.inductiveautomation.com/u/martin.johansson)\
**Post date:** [August 17, 2020, 1:02pm UTC](https://forum.inductiveautomation.com/t/perspective-alarm-journal-causes-100-cpu/37804/1 "2020-08-17T13:02:43Z")

</div>

Hi!

We are trying out Ignition to store and show alarms. Been adding a few production-lines so far and have around 2000 configured alarms. Mostly simple booleans imported from Kepware together with a description tag.

It works well, but when you select a long time span and the journal shows ~100.000 alarms the gateway maxes out in CPU use and does that till you close the journal. I have tried to make a new project and only adding an Alarm Journal with the same behavior. I would expect that it takes some time to load all entries but then the load should stop, shouldn’t it?

When I try to use Vision instead everything works as expected. It takes a minute or so to load but when it’s done the CPU goes back to idle.

Is this a known issue, and is there something I can do to fix it?

---

<div class="post-metadata">

**Author:** ![MMaynard](https://sea1.discourse-cdn.com/inductiveautomation/user_avatar/forum.inductiveautomation.com/mmaynard/32/3366_2.png) [@MMaynard](https://forum.inductiveautomation.com/u/MMaynard)\
**Post date:** [August 17, 2020, 5:01pm UTC](https://forum.inductiveautomation.com/t/perspective-alarm-journal-causes-100-cpu/37804/2 "2020-08-17T17:01:30Z")

</div>

I would start by adding indexes on the alarm table(s) columns.  
We normally add Ascending/Descending indexes on the eventtime columns in alarm\_events  
We also have stored procedures and do our own queries to display historical alarms and only use the journal to display active alarms.

---

<div class="post-metadata">

**Author:** ![martin.johansson](https://avatars.discourse-cdn.com/v4/letter/m/7feea3/32.png) [@martin.johansson](https://forum.inductiveautomation.com/u/martin.johansson)\
**Post date:** [August 18, 2020, 7:52am UTC](https://forum.inductiveautomation.com/t/perspective-alarm-journal-causes-100-cpu/37804/3 "2020-08-18T07:52:44Z")

</div>

Thanks for your replay. I did add some indexes and it might have made the initial search a bit faster but the main problem still remains.

I opened the Alarm Journal and listed 24weeks back, around 170000 alarms. And then the CPU is stuck at ~100%. This time I left the journal open during breakfast and when I came back it’s still on ~100%. Closed the browser and still, the CPU continues in high load, even after half an hour.

![Skärmklipp](https://us1.discourse-cdn.com/inductiveautomation/original/3X/9/c/9cbdae0d059866a910adeaf28ed9b1e5c039a361.png)

If I check the threads perspective-scheduler stays around 30%. And if I check my session total bytes sent says a whopping 2GB.

Viewing history with your own queries might be a solution as those seems to work fine. But I would still like to use the Journal as it’s a finished product, and works perfectly fine in Vision.

EDIT: Zulu Platform x64 is stuck at 100% and if I force quit it, it just restarts.