Perfbase database update to version 5

View: New views
2 Messages — Rating Filter:   Alert me  

Perfbase database update to version 5

by Sagi Rotem :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

Joachim hi,

we had updated the database for the svn and tried to upgrade the old data to the new database format.

I am getting the following errors :

trying to upgrade all the database :

/mswg/mpi_db>perfbase check -a
FATAL:  database "pb_perfbase" does not exist

trying to upgrade the pmb (pallas) experiment

/mswg/mpi_db>perfbase check -e pallas -u
Traceback (most recent call last):
  File "/mswg/mpi_db/bin/pb_check.py", line 590, in ?
    main()
  File "/mswg/mpi_db/bin/pb_check.py", line 575, in main
    update_experiment()
  File "/mswg/mpi_db/bin/pb_check.py", line 327, in update_experiment
    raise SpecificationError, "corrupted experiment '%s': no metadata stored." % exp_name
pb_common.SpecificationError: "corrupted experiment 'pallas': no metadata stored."

Is this the right way to upgrade the database ?

Thanks,

Sagi


Re: Perfbase database update to version 5

by Joachim Worringen :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

Sagi Rotem wrote:

> Joachim hi,
>
> we had updated the database for the svn and tried to upgrade the old
> data to the new database format.
>
> I am getting the following errors :
>
> trying to upgrade all the database :
>
> /mswg/mpi_db>perfbase check -a
> FATAL:  database "pb_perfbase" does not exist

Hi Sagi,

this is a bug with the -a switch witch did not work any longer. I have fixed
this problem and committed this version.

> trying to upgrade the pmb (pallas) experiment
>
> /mswg/mpi_db>perfbase check -e pallas -u
> Traceback (most recent call last):
>   File "/mswg/mpi_db/bin/pb_check.py", line 590, in ?
>     main()
>   File "/mswg/mpi_db/bin/pb_check.py", line 575, in main
>     update_experiment()
>   File "/mswg/mpi_db/bin/pb_check.py", line 327, in update_experiment
>     raise SpecificationError, "corrupted experiment '%s': no metadata
> stored." % exp_name
> pb_common.SpecificationError: "corrupted experiment 'pallas': no
> metadata stored."
>
> Is this the right way to upgrade the database ?

If you experiment is named 'pallas', it is the right way. But I guess your
experiment is named 'Pallas'? This name is case sensitive. The exception that is
generated should be catched - the fixed version does this corrrectly.

  Joachim

--
Joachim Worringen - NEC C&C research lab St.Augustin
fon +49-2241-9252.20 - fax .99 - http://www.ccrl-nece.de

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@...
For additional commands, e-mail: users-help@...