Scheduled Meta Sync [Jul 08]

Discord-Discussions
1

[dummy title body text appended to avoid empty title body]

2

:raising_hand_man: amarnath replied

still the same, cant get

3

:raising_hand_man: amarnath replied

when i tried deploying in ecs too, i have the same problem

4

:raising_hand_man: amarnath replied

health check for elb is failing

5

:raising_hand_man: wkw replied

can u remove this line -e NC_DB="mysql2://host.docker.internal:3306?u=root&p=password&d=d1" \ n try again

6

:raising_hand_man: wkw replied

if it works, then it may be sth wrong with your mysql instance

7

:raising_hand_man: wkw replied

for elb failing, it s hard to tell, you may need to check

- the connection from elb to ecs

- the connection between ecs n db

i.e. like security groups - make sure the ports are opened n the traffic are allowed etc

8

:raising_hand_man: wkw replied

cuz we got some users using ECS and it is working fine

9

:raising_hand_man: amarnath replied

i am using terraform to control all the changes. my present nocodb is 0.84.14(task definition 15), and trying to upgrade to 0.91.1 (task definition 16)

while 15 is working, 16 is not.

and i have removed NC_DB, now its working locally

10

:raising_hand_man: amarnath replied

```➜ ~ diff task1 task2

1d0

<

67c66

< “image”: “nocodb/nocodb:0.84.14”,

—

> “image”: “nocodb/nocodb:0.91.1”,

94c93

< “taskDefinitionArn”: “arn:aws:ecs:ap-southeast-1:33333914:task-definition/nocodb:15”,

—

> “taskDefinitionArn”: “arn:aws:ecs:ap-southeast-1:30333914:task-definition/nocodb:16”,

135c134

< “revision”: 15,

—

> “revision”: 16,```

11

:raising_hand_man: amarnath replied

diff of my task definition

12

:raising_hand_man: amarnath replied

any idea what should be the issue?

13

:raising_hand_man: wkw replied

that would be lots of changes in NC_DB for the upgrade to 0.90.X

14

:raising_hand_man: wkw replied

if you cant get into dashboard, then it means something is wrong with your NC_DB (as you proved that by not using NC_DB)

15

:raising_hand_man: wkw replied

so now i think you only got 2 choices

1. use a new DB as NC_DB (this should be faster if all of your projects are connecting to external DB but you have to create those projects again and meta will be lost such as sorting / filtering etc)

2. otherwise, u need to check Cloudwatch logs to figure out what goes wrong with the upgrade

16

:raising_hand_man: amarnath replied

1st one is not an option. there are lot of projects. why 84.14 is working, 0.90.0 is not (i have tested this version as well)

17

:raising_hand_man: amarnath replied

both in my local and ecs

18

:raising_hand_man: wkw replied

bc we introduced some breaking changes (we revised the entire API design)

19

:raising_hand_man: wkw replied

which is mentioned in v0.90 Important Announcement · Issue #1564 · nocodb/nocodb · GitHub

20

:raising_hand_man: wkw replied

> both in my local and ecs

it doesnt relate to your local or ecs. it is related to your NC_DB as we would try to convert the old meta to our new design. and it seems something went wrong but we need the details.