https://signoz.io
Join Slack
This message was deleted.
# general
s
This message was deleted.
✅ 1
a
I look at trace details then it all goes black and blank.
@Robert Smith this used to occur when spans are missing and a complete trace is not built. In 
v0.10.0
we have handled this and blank screen does not appear anymore..rather it displays set of spans that are still pending for trace to get completed
which version of signoz are you running?
r
This is a first install and I'm running 3.9 according to the docker-compose.yml file
Ah, having got the frontend running again I see the actual Signoz version I'm running is 0.10.0
p
got it - can you share a screenshot of what you are seeing @Robert Smith Would help us understand the issue better
a
screenshot and the response of API call at that page would be very helpful. Response of api call can be found at network tab after inspect element
r
These show the traces screen I was in and then the blank one after clicking any of the traces shown. After about 1/2 second some stuff did flash up on the screen but too quickly for me to register any content the screen turned black as show. This behaviour is repeatable and happens every time for any trace I select.
I'm not sure how to access the API call info as my screen is black showing nothing at that point. I have clicked 'refresh' a few time to read the message flashing up and it is as follows:
A) Screen shows a blue 'loading' graphic in centre of screen for 1/4 second, on just one occasion for 2 seconds. B) The lefthand menu pane appears along with a black main pane showing at the top only Home/Trace/ and then some long number in white. Then it immediately goes completely black for the whole window.
a
which system are you running signoz on? I see you are on latest version 
v0.10.0
on bottom-left corner but it should be a green tick?
r
Running on a VM in Oracle OCI cloud running OrackleLinux/Redhat 8.6
a
ohk...do you see any browser console errors in that page?
Running on a VM in Oracle OCI cloud running OrackleLinux/Redhat 8.6
@Prashant Shahi any idea if this is causing some issue?
r
Funnily enough when I went back to a non-blank prior screen there was a green tick, but 5 seconds later that cjhanged back to the red triangle. (P.S. I had a quick go at installing Loom but failed on that - our windows admins do stop us trying to install anything on the desktops).
I just went hunting for log files and found just those in signoz/deploy/docker/clickhouse-setup/data/clickhouse/data/system but when I went there they were all links to stuff in /var/lib/clickhouse/store/ but there is no such location so the links are all bad. This doesn't seem good ...
foobar and time to uninstall and reinstall?
p
we haven't tested in Oracle Linux.. but it shouldn't be a problem
Can you share the steps you followed to set up SigNoz cluster?
r
Sure. It is just a single node set up for a POC where I have just followed the online docker-standalone install process as here https://signoz.io/docs/install/docker/ where I didn't choose a specific version but installed from main.
ah, didn't mean to have that screenshot showing. sorry
This may be my fault. Reading troubleshooting & its links to pre-reqs I get the idea that trying to run this on a small 1G VM is not going to fly. Is the suggested 4G memory sufficient for small scale trial use or, as I'll have to provision another VM, should I allocate more?
a
yes.. 1G can show unexpected errors
2CPUs and 4GB RAM for a PoC and scaling as needed
r
OK, thanks for your help. I'll try turning the metric retention down to see if that makes it usable first then warn the bean counters that I'm about to spnd some money.
a
I'll try turning the metric retention down
metrics do not have any relation to trace detail page where you are seeing errors
r
Both turning metric retention down and increasing existing VM size to 2xcpu and 16G ram has not changed the behaviour. Still not finding the trace and giving a blank screen - might this be something to do with missing /var/lib/clickhouse/store/? Anything left to do now rather than uninstall and reinstall, and if so should I install from main or a specific version?
Deleting and reinstalling everything from scratch still hasn't changed the problem. Unless this is a reproducible problem elsewhere with current main version I'm tempted to go back to my original idea that it might be the very old version of Chrome that we have installed on the desktops.
Resolved: Thanks for the help guys. Problem was in the end the old browser so as a data point Chrome v79.x fails but Chrome v88.x seems fine.
🎉 2
p
awesome @Robert Smith Looking forward to learn more on how you are using SigNoz and any feedback/suggestions 🙂
15 Views
Open in Slack
PreviousNext
Powered by

SigNoz Community

SigNoz is an open-source APM. It helps developers monitor their applications & troubleshoot problems, an open-source alternative to DataDog, NewRelic, etc.

Powered by