https://signoz.io logo
Join Slack
Powered by
Hi team, I am migrating from 0.76 to 0.88 (Docker ...
# support
e
Hi team, I am migrating from 0.76 to 0.88 (Docker standalone). I setup both exporters, failed at migrating data (but not needed) , have issues migrating alerts and dashboard. Thanks for your help in advance! 🙂
s
e
the attributes map was added, now I get a different error with tables?
s
Ensure the db path is correct, @Nagesh Bansal help here
e
this db path?
s
correct, the error says no table 
dashboard
that shouldn't be the case.
I assume you were already on .76 and then migrating to 0.88 https://signoz.io/docs/operate/upgrade/#upgrade-path
e
yes, I am following the instructions of upgrade-0.88
s
volume mapped correctly?
e
I am in 0.76, doing the instructions before upgrading to 0.88
image.png
I am not so sure where should be the sqlite file in the drive
the signoz.db file exists
I was using v0.72 , then migrated to 0.76 last week and now migrating to 0.88. We cannot use 0.76 because as soon I docker compose up , cpu goes high and it is unusable (I hope migrating helps with that issue)
s
Can you exec into and check the contents of signoz.db
e
yes, looks like this
there is a part that says dashboards
well, the error points to table "dashboard" but in db it says "dashboards"
s
It was changed some time back @Nagesh Bansal knowns more details
n
Hey @Efrain Sotelo Ferry, can you please first try to upgrade to 0.86.2 and then try with the dot metrics migration, i.e 0.88.0
e
Hi @Nagesh Bansal , I will try that tomorrow, today I am not at work.
👍 1
Hi @Nagesh Bansal,I manage to upgrade my alerts and alarms to 0.86.2 and then to 0.88.0. But when my services starts sending to the collector the instance of signoz gets unresponsive (CPU% increases a lot). Can you guide on steps to debug this issue?
n
Ohh Great!! It might be because the dual exporters are being enabled. Please monitor your system for a few days to ensure stability, then you can disable clickhousemetricswrite exporter and upgrade to the latest version ( as per your needs)
e
it is already disabled, I went through and upgraded to 0.90.1
👍 1
the cpu increase was the reason I also started the migration, probably it is not due to the migration itself
I am restarting the instance and preparing to only receive metrics from one service and being able to see the docker compose logs
which service or what should I check for this CPU increase anomaly?
n
@Srikanth Chekuri, do you have any idea why they might be experiencing high CPU usage?
e
I rebooted instance, up signoz(docker compose up -d) but I block port 4317 so no logs reaching to otel-collector, I did this to check logs and found a log from clickhouse that says error:
@Nagesh Bansal @Srikanth Chekuri my clickhouse docker volume is 70 GB now, would this be a possible issue?, if anyways I only migrated alerts and dashboard, is it safe to reset the volume and start from scratch?
s
is it safe to reset the volume and start from scratch?
Do you mean ClickHouse volume? I would suggest truncate instead.
e
yes , clickhouse volume. Truncate https://clickhouse.com/docs/managing-data/truncate .
@Srikanth Chekuri last question if I may, when I upgrade signoz , the signoz-otel-collector container can also be updated , f.e. 0.128.2. My services uses a sidecar otel-collector and the sidecar sends logs to the signoz-otel-collector. Do I have to update the sidecar otel-collector version? if yes, how to know which version is compatible with signoz otel-collector?
s
Do I have to update the sidecar otel-collector version?
No
e
@Srikanth Chekuri @Nagesh Bansal thanks for your help 🙌
🙌 1
5 Views
Open in Slack
PreviousNext
Powered by