https://linen.dev logo
Join Slack
Powered by
Hi there! I’m trying to get a meltano project set ...
# troubleshooting
b
Hi there! I’m trying to get a meltano project set up, and receiving an odd error message:
Copy code
uvicorn@ea132a6a9580:/opt/wg$ meltano install
Installing 1 plugins...
Installing loader 'target-postgres'...
'coroutine' object has no attribute 'read'
sys:1: RuntimeWarning: coroutine 'StreamReader.read' was never awaited
RuntimeWarning: Enable tracemalloc to get the object allocation traceback
uvicorn@ea132a6a9580:/opt/wg$ pip freeze | grep meltano
meltano==1.73.0
uvicorn@ea132a6a9580:/opt/wg$ cat meltano.yml 
version: 1
send_anonymous_usage_stats: false
plugins:
  loaders:
  - name: target-postgres
    variant: datamill-co
    pip_url: singer-target-postgres
I’ve tried version 1.75.0 as well, and same result
This is python 3.8.5 on a ubuntu 20.04 image in docker
d
Can you try 
pip install --upgrade pip
before 
meltano install
? I always get errors unless I do this (maybe not this particular one though)
b
pip was already up to date (it’s in our Dockerfile). anyone else have any thoughts?
v
The error message isn't helpful here (issue after you find it would be helpful I think), but your pip_url looks wrong to me.
It's not wrong (I was wrong about the pipurl)
https://hub.meltano.com/targets/postgres I ran through these steps and it did work for me. My gut is
Copy code
'coroutine' object has no attribute 'read'
Is trying to tell you there's an access problem or something, but that's a complete guess. Running this in debug mode may be helpful 
meltano --log-level=debug install
d
I wonder if this is coming from https://gitlab.com/meltano/meltano/-/commit/e3655e7f4deb9db0dbbfcdc2156a8beb6a134790#3e6829026c98d73166d80cafa395dc62469b6d8e_38_52, which was introduced in https://gitlab.com/meltano/meltano/-/merge_requests/2104. @beau_cronin Can you please file an issue so we can investigate further?
b
That was my basic guess once I saw that parallel installs were a new thing, and that the error was async io-related. That’s why I went back to 1.73, with the understanding that that version did not have that new functionality?
In any case, I will file an issue.
Also @douwe_maan - if that is the problem, what might be a short-term workaround? Need to keep this work moving forward in the meantime if at all possible
d
Ah good call, if you're also seeing this in 1.73.0, it's probably not that feature. I missed that 🙂
With just what we're seeing in these logs, it's hard to identify a workaround. I suggest joining office hours; we may have some time at the end to debug this live: https://meltano.slack.com/archives/C01TCRBBJD7/p1623859414262700
@beau_cronin Did you go back to 1.72.0? Parallel install was introduced in 1.73.0: https://gitlab.com/meltano/meltano/blob/master/CHANGELOG.md#1730-2021-04-29
b
ah, I might have not gone back quite far enough. let me try that and I’ll report back
ok, update: • 1.72.0 requires pandas/numpy versions for which wheels are not distributed • Avoiding those builds would mean going to python3.7, which is not easy on Ubuntu 20.04 (and we use python3.8 everywhere else, so not a good fix anyway) • So afaict 1.72.0 on Ubuntu 20.04 is a tough combo
d
Hmm, that's annoying
@beau_cronin Could you temporarily run 1.72.0 via Docker? See https://meltano.com/docs/getting-started.html#docker-installation. 
docker pull meltano/meltano:v1.72.0
The fact that 
meltano install
is even getting to this line: https://gitlab.com/meltano/meltano/-/commit/e3655e7f4deb9db0dbbfcdc2156a8beb6a134790#3e6829026c98d73166d80cafa395dc62469b6d8e_38_52, suggests that 
pip install target-postgres
failed for some reason, so maybe you can run that by itself and figure out what the error is. If you address the issue, 
meltano install
may just succeed without getting to that broken code path
An issue to fix the actual bug would still be appreciated 🙂
b
yes, I will def file an issue regardless - will need to wait until later today
thanks for the help so far, and I’ll keep poking around between other things today
3 Views
Open in Slack
PreviousNext
Powered by