http://coldfusion.com logo
Join SlackCommunities
Powered by
How do I prevent FusionReactor from losing histori...
# fusion-reactor
t
How do I prevent FusionReactor from losing historical data after it's restarted?
b
@Tim Badolato Use the "archived metrics" menu to find old data
There is no way to prevent the in-memory data from being lost which is kept since the last restart
So stuff like "longest requests" or "errored requests" is gone, but the overall request log is still there in archived metrics as well as most of the memory/CPU/traffic sort of graphs
The main limiting factor is the data is carved up into 1 hour segments. (the logs presumably rotate hourly) so you need to have a basic idea of the time in history you want to look at
t
The issue I'm running into is a long running request that kills the server. And once the server restarts, I can't determine which request caused it or what the cause was.
b
Use the E-mail feature
You can configure alerts based on any number of triggers including request time
t
The email feature will email the request that caused the crash?
b
It sends you a list of all running requests, all runing queries, and a thread dump of the entire JVM
It's your go-to tool for post-mortem questions like "what was running when the server backed up"
You can play with the settings (how many concurrent requests are running, longest request time, etc) to get useful E-mails without spamming your inbox
The feature also won't E-mail you more than X times per minute, etc
t
Nice! I definitely need to get that setup. Thankfully the crash is extremely rare and usually during a few peak times of the year. I'm just curious if I can optimize things to prevent it.
b
Yep, all you need is to set your mail server settings
🎉 1
Do that in the menu you get clicking the fusionreactor logo in the very top left of the screen and select "settings"
image.png
Click "send test E-mail" to check the settings
⭐ 1
d
@bdw429s Once again, you're super handy to have around!
😁 1
b
Then go to "protection" > "settings"
image.png
Depending on what your symptoms are (long running requests, too many requests stacking up, high CPU, etc) you can choose the tab you want and set the drop down to "E-mail notification only"
How often it pings you can be configured under "E-mail settings" at the bottom of the sub nav
t
Do you leave all the other Crash Protections settings on default?
b
Usually. Only the rules you activate will do anything
I don't ever tend to have FR try to kill requests. Hung requests rarely die anyway.
👍 1
d
Just to say it, and I could be wrong since we can't do this for PHI reasons, I think this is only an issue with on-premise FR. The cloud version should have everything across restarts. I think.
b
Correct, FR Cloud retains all information, but it also has much less info. Only "transactions of interesting" (long running, erroring, etc) are ever sent to FR cloud
👍 1
2 Views
Open in Slack
PreviousNext
Powered by