deploy.sh seems broken on a Clearlinux host

[clsulliv]

Running the following:

./deploy.sh -H localhost -d clr -a install

Results in:

Command "/var/www/telemetry/venv/bin/python3 -u -c "import setuptools, tokenize;__file__='/tmp/pip-install-7ceum3p0/psycopg2/setup.py';f=getattr(tokenize, 'open', open)(__file__);code=f.read().replace('\r\n', '\n');f.close();exec(compile(code, __file__, 'exec'))" install --record /tmp/pip-record-hnqv6vfg/install-record.txt --single-version-externally-managed --compile --install-headers /var/www/telemetry/venv/include/site/python3.7/psycopg2" failed with error code 1 in /tmp/pip-install-7ceum3p0/psycopg2/

This can be resolved by upgrading to psycopg2==2.7.6 in the deploy.sh script.

After that the installation completes successfully, but the web UI doesn't come up, and instead I get a 500 server error. The journal shows many issues like the following:

Jan 08 13:13:30 clsulliv-nuc postgres[24208]: ERROR: column records.os_name does not exist at character 8
Jan 08 13:13:30 clsulliv-nuc postgres[24208]: STATEMENT: SELECT records.os_name AS records_os_name, build.build AS build_build
Jan 08 13:13:30 clsulliv-nuc postgres[24208]: FROM records JOIN build ON build.id = records.build_id GROUP BY records.os_name, build.build ORDER BY records.os_name

After doing some digging I found that I can resolve this by downgrading the database via the following from the telemetryui folder:

sudo bash -c "export https_proxy=$https_proxy; source ../venv/bin/activate && FLASK_APP=run.py flask db downgrade"

It seems that the web ui are database are mismatched somehow.

[tpleavitt]

@alexjch , I've run into this same build error as well. The error message I get when I run the browser for "localhost" is "internal server issue", not a 500 error. Also, the script calls out psycopg2==2.7.6 (mentioned by @clsulliv above) and successfully gets it.