Troubleshooting an Upgrade
Problem
Seeq server fails to start after an upgrade. The “Launch Seeq” UI may display a message saying Seeq has failed to start and Seeq support should be contacted.
Solution
Insufficient file permissions on the Seeq data folder and installation directory can cause a myriad of problems. Try loosening file permissions by executing the following command from the Seeq Command Line Interface (CLI) to give Everyone permissions on the Seeq data folder and installation directory:
seeq diag fix -f
After the command completes, restart Seeq. If the “Launch Seeq” UI shows that the upgrade fails to progress or Seeq still cannot start, please contact Seeq support for further assistance.
Problem
Seeq does not start after migrating a data folder from one server to another; additionally, the log file <data>/log/pilot/postgres.log has a pattern of repeated attempts to start postgres that error out, like so:
DEBUG 2021-03-12T14:19:48.603 --------------------------------------------------------------------------------
DEBUG 2021-03-12T14:19:48.604 seeq db start -t postgres --disable-keypress
DEBUG 2021-03-12T14:19:48.604 --------------------------------------------------------------------------------
...
ERROR 2021-03-12T14:19:49.139 ***** ERROR *****
ERROR 2021-03-12T14:19:49.139 Command failed with exit code 1: "C:\Program Files\Seeq Server\postgres\files\bin\pg_ctl.exe" start -t 21600 -D "C:\ProgramData\Seeq\data\postgres"
If the command seeq db start -t postgres --disable-keypress
is run at the Seeq Command Prompt after shutting down, the database starts up without issue.
Solution
Check and adjust the memory settings for Postgres. If the host to which the data folder has been migrated does not have the same memory as the previous host (this is common in a dev/prod deployment scenario), there may not be enough memory for Postgres to start up. In particular, Memory/Postgres/Size
, Database/Postgres/WorkMem
, and Database/Postgres/SharedBuffers
, can cause startup issues or subsequent problems if they were overridden on the source host and not adjusted upon migration to the new host.
Problem
If Seeq doesn’t start in interactive mode and you see the following messages in the unified.log
you might have an issue with the account you are logged in.
[AuthPostgres ] "D:\ApplProd\Seeq\Seeq Server\postgres\files\bin\initdb.exe" -A scram-sha-256 -D "D:\ApplProd\Seeq\Data\auth\postgres-initdb-tmp" -U seeq --pwfile "D:\ApplProd\Seeq\Data\auth\keys\auth_service.key" -E UTF8 --locale C
[AuthPostgres ] ***** ERROR *****
[AuthPostgres ] Command failed with exit code 3221225781: "D:\ApplProd\Seeq\Seeq Server\postgres\files\bin\initdb.exe" -A scram-sha-256 -D "D:\ApplProd\Seeq\Data\auth\postgres-initdb-tmp" -U seeq --pwfile "D:\ApplProd\Seeq\Data\auth\keys\auth_service.key" -E UTF8 --locale C
[AuthPostgres ] "D:\ApplProd\Seeq\Data/log/pilot/postgres-auth.log" may have more information.
[Monitor ] Starting up process monitor
[Monitor ] Crashed with unexpected error: int() argument must be a string, a bytes-like object or a number, not 'NoneType'
[SeriesCachePostgres ] "D:\ApplProd\Seeq\Seeq Server\postgres\files\bin\initdb.exe" -A scram-sha-256 -D "D:\ApplProd\Seeq\Data\cache\series\postgres-initdb-tmp" -U seeq --pwfile "D:\ApplProd\Seeq\Data\cache\series\keys\cache.key" -E UTF8 --locale C
[SeriesCachePostgres ] ***** ERROR *****
[SeriesCachePostgres ] Command failed with exit code 3221225781: "D:\ApplProd\Seeq\Seeq Server\postgres\files\bin\initdb.exe" -A scram-sha-256 -D "D:\ApplProd\Seeq\Data\cache\series\postgres-initdb-tmp" -U seeq --pwfile "D:\ApplProd\Seeq\Data\cache\series\keys\cache.key" -E UTF8 --locale C
[SeriesCachePostgres ] "D:\ApplProd\Seeq\Data/log/pilot/series-cache-postgres.log" may have more information.
[PostgreSQL ] "D:\ApplProd\Seeq\Seeq Server\postgres\files\bin\initdb.exe" -A scram-sha-256 -D "D:\ApplProd\Seeq\Data\postgres-initdb-tmp" -U seeq --pwfile "D:\ApplProd\Seeq\Data\keys\postgres_admin.key" -E UTF8 --locale C
[PostgreSQL ] ***** ERROR *****
[PostgreSQL ] "D:\ApplProd\Seeq\Data/log/pilot/postgres.log" may have more information.
[PostgreSQL ] Command failed with exit code 3221225781: "D:\ApplProd\Seeq\Seeq Server\postgres\files\bin\initdb.exe" -A scram-sha-256 -D "D:\ApplProd\Seeq\Data\postgres-initdb-tmp" -U seeq --pwfile "D:\ApplProd\Seeq\Data\keys\postgres_admin.key" -E UTF8 --locale C
Solution
Try a full admin account or run Seeq as a service using the Local System account.
