Hi,
I've been onboard for several years without upgrade issue but hit a wall when moving from 52.17 to 53.1. IA64 jar style updates using NSSM as a service tool. this has been rock solid for me.. In this upgrade case I watch the cpu utilization and it never settles down as usual with an upgrade.
Was there anything in particular with this update I should know?
Thanks,
Regis
1G is extremely low, use 2-3G at least
Did that, -Xmg3g (its a 16gb system memory). still goes into infinite cpu utilization loop. Is there a log that can be reviewed during the upgrade? If you can provide where to access it. Thank you.
Regis
Send the application logs when the spike starts. I don’t know how is the logging configured in the windows machine
I cannot get Metabase running to see the logs. Can you please refer me to someone familiar with x64 architecture where i can see the logs on disk? I tried from java help but no success.
@crwheelr when Metabase starts, it outputs the logs, so you should have the logs in the service console in windows, a text file or something. This is not about x64, amd64, risc or anything related, you need to find the person who configured Metabase in that windows server to know where the logs are being streamed to and send those to pinpoint the issue
I installed over 5 years ago and have done countless upgrades by stopping the service with NSSM, replacing the exe and restarting it. never a problem. i noticed that the .exe size went from 421mb to 473mb for this version jump i am having problems with . i have 3gb allocated. is there anything specific with this release that might have me in a continuous loop of cpu hogging during first startup?
i didn't do any configuration regarding logs. where is the default windows /metabase log?
thx
Regis
@luiggi, i checked windows event viewer and dont see anything under applications/Metabase. If you are not sure, can you please fwd this request of logging? thx - regis
@crwheelr just to align expectations here: there's no other person or team to forward this to, and also we can't tell you how to fix a deployment issue if we can't see the logs that the application spits when it runs, so unless you can find that it's a dead end and we would recommend that you use any of the recommended tutorials to deploy Metabase that we have on the website
@luiggi, Okay I can see that. I will comb through the deployment documentations to find the windows logging file and location. Then I will be able to get you something more meaningful. - regis
@luiggi, created the log files but cannot attach them here (only picture files it seems). Is there somewhere I can attach?
I have also created refined batch files for nssm install with logging I'd like to share with others.
Here are the first lines of the error log with v0.53.1
Exception in thread "main" java.lang.NoClassDefFoundError: java/util/SequencedCollection
at java.base/java.lang.Class.forName0(Native Method)
at java.base/java.lang.Class.forName(Unknown Source)
at clojure.lang.RT.classForName(RT.java:2229)
at clojure.lang.RT.classForName(RT.java:2238)
at instaparse.auto_flatten_seq$fn__6961.(auto_flatten_seq.clj:314)
at instaparse.auto_flatten_seq__init.load(Unknown Source)
at instaparse.auto_flatten_seq__init.(Unknown Source)
at java.base/java.lang.Class.forName0(Native Method)
at java.base/java.lang.Class.forName(Unknown Source)
v0.52.17 has no error log content
is there a way you can create a public google drive link where you put the log files? the specific lines you posted are a great progress, but we would like to have the logs since Metabase starts till the moment you get the blank screen
@Luiggi , have you any time to study the logs? thx, Regis
