Labscript suite installation issues

[Pierre Colombo]

Hello,

Thanks for your work on Labscript.

I meet problems when I install Labscript, i have followed the https://bitbucket.org/labscript_suite/installer.

Configuration :

- Windows 64 bits

- Anaconda 2.7

- zmq version 4.1.5 

-zprocess version 1.3.3

-pyzmq version 15.4.0

When I try to launch runviewer i get :

Traceback (most recent call last):

  File "C:\labscript_suite\runviewer\__main__.py", line 47, in <module>

    import zprocess.locking, labscript_utils.h5_lock, h5py

  File "C:\labscript_suite\labscript_utils\h5_lock.py", line 98, in <module>

    connect_to_zlock_server()

  File "C:\labscript_suite\labscript_utils\h5_lock.py", line 89, in connect_to_zlock_server

    zprocess.locking.connect(host,port,timeout=15)

  File "C:\Program Files\Anaconda2\lib\site-packages\zprocess\locking\__init__.py", line 373, in connect

    ping(timeout)

  File "C:\Program Files\Anaconda2\lib\site-packages\zprocess\locking\__init__.py", line 337, in ping

    return _zmq_lock_client.say_hello(timeout)

  File "C:\Program Files\Anaconda2\lib\site-packages\zprocess\locking\__init__.py", line 114, in say_hello

    raise zmq.ZMQError('No response from zlock server: timed out')

ZMQError: No response from zlock server: timed out

Runmanager :

Traceback (most recent call last):

  File "C:\labscript_suite\runmanager\__main__.py", line 66, in <module>

    import runmanager

  File "C:\labscript_suite\runmanager\__init__.py", line 25, in <module>

    import labscript_utils.h5_lock

  File "C:\labscript_suite\labscript_utils\h5_lock.py", line 98, in <module>

    connect_to_zlock_server()

  File "C:\labscript_suite\labscript_utils\h5_lock.py", line 89, in connect_to_zlock_server

    zprocess.locking.connect(host,port,timeout=15)

  File "C:\Program Files\Anaconda2\lib\site-packages\zprocess\locking\__init__.py", line 373, in connect

    ping(timeout)

  File "C:\Program Files\Anaconda2\lib\site-packages\zprocess\locking\__init__.py", line 337, in ping

    return _zmq_lock_client.say_hello(timeout)

  File "C:\Program Files\Anaconda2\lib\site-packages\zprocess\locking\__init__.py", line 114, in say_hello

    raise zmq.ZMQError('No response from zlock server: timed out')

ZMQError: No response from zlock server: timed out

Lyse :

Traceback (most recent call last):

  File "C:\labscript_suite\lyse\__main__.py", line 23, in <module>

    import labscript_utils.h5_lock

  File "C:\labscript_suite\labscript_utils\h5_lock.py", line 98, in <module>

    connect_to_zlock_server()

  File "C:\labscript_suite\labscript_utils\h5_lock.py", line 89, in connect_to_zlock_server

    zprocess.locking.connect(host,port,timeout=15)

  File "C:\Program Files\Anaconda2\lib\site-packages\zprocess\locking\__init__.py", line 373, in connect

    ping(timeout)

  File "C:\Program Files\Anaconda2\lib\site-packages\zprocess\locking\__init__.py", line 337, in ping

    return _zmq_lock_client.say_hello(timeout)

  File "C:\Program Files\Anaconda2\lib\site-packages\zprocess\locking\__init__.py", line 114, in say_hello

    raise zmq.ZMQError('No response from zlock server: timed out')

ZMQError: No response from zlock server: timed out

And blacs doesn't seem to work .

Do you have any ideas on how to correct this server connection pb ?

Best

[Chris Billington]

Hi Pierre,

This is due to a faulty version of the zprocess module that I accidentally published. I thought I had taken it down but clearly I have not done so correctly. I apologise. You can downgrade to the unbroken version of zprocess with:

pip install zprocess==1.3.3

From the Windows command line. (you may need to type the full path to pip, 

C:\Program Files\Anaconda2\bin\pip, if this doesn't work).

Regards, 

Chris

--
You received this message because you are subscribed to the Google Groups "The labscript suite" group.
To unsubscribe from this group and stop receiving emails from it, send an email to labscriptsuite+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Pierre Colombo]

Hi Chris,

Thanks for your fast reply in fact i have installed zprocess 1.3.3 as you suggest in the previous post .

Chris

To unsubscribe from this group and stop receiving emails from it, send an email to labscriptsuit...@googlegroups.com.

[Chris Billington]

Hi Pierre,

My mistake, you did mention the version numbers of everything, thanks for that.

In that case, we should verify whether "zlock" is configured correctly. zlock is a server program that all other programs in the labscript suite use to coordinate their access to files, so that they don't simultaneously write to the same files as each other. A zlock server must be running in order for the other programs to connect to it, and what you're seeing here are errors saying that the programs could not connect to the zlock server.

Could you let me know what your configuration in C:\labscript_suite\labconfig\<your_computers_hostname>.ini is? There should be a section called [servers], could you let me know what the zlock server is set to?

By default, it is set to localhost - so it shouldn't have changed unless you changed it yourself. If it set to localhost, then programs that need to use zlock should launch a server automatically if one is not already running. However, if it's set to some other computer, then a zlock server needs to be manually started on that computer.

A zlock server can be started manually with:

python -m zprocess.locking

from the command line. Running this line should give you an error if a zlock server is already running, which could be useful to know.

Regards,

Chris

To unsubscribe from this group and stop receiving emails from it, send an email to labscriptsuite+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

Chris Billington
School of Physics and Astronomy
Monash University
Victoria 3800, Australia
Ph: +61 423952456

[Pierre Colombo]

Hi Chris,

This is the content of the file : server seems correct,

[DEFAULT]

experiment_name = example_experiment

shared_drive = C:

experiment_shot_storage = %(shared_drive)s\Experiments\%(experiment_name)s

labscript_suite = C:\labscript_suite

labscriptlib = %(labscript_suite)s\userlib\labscriptlib\%(experiment_name)s

analysislib = %(labscript_suite)s\userlib\analysislib\%(experiment_name)s

pythonlib = %(labscript_suite)s\userlib\pythonlib

[servers]

zlock = localhost

[ports]

blacs = 42517

lyse = 42519

mise = 42520

runviewer = 42521

zlock = 7339

[programs]

text_editor = C:\Program Files (x86)\Notepad++\notepad++.exe

text_editor_arguments = {file}

hdf5_viewer = C:\Program Files\TheHDFGroup\HDFView2.7\HDFView.exe

hdf5_viewer_arguments = {file}

[paths]

connection_table_h5 = %(experiment_shot_storage)s\connectiontable.h5

connection_table_py = %(labscriptlib)s\connectiontable.py

[runmanager]

autoload_config_file = %(experiment_shot_storage)s\runmanager.ini

output_folder_format = %%Y\%%m\%%d

[lyse]

autoload_config_file = %(experiment_shot_storage)s\lyse.ini

If I try to start the server manually I get : zmq.error.ZMQError : Adress in use, I guess this means that the server is runing ?

But this is not working.

(please find attached the screenshot)

Thanks for your time,

best,

Pierre

[Chris Billington]

Thank you,

Well, this looks very much like the bug in zprocess 2.0.0. Did you install zprocess 2.0.0 at some point, and then downgrade? If so, it would be good to confirm zprocess 2.0.0 is really gone, which you could o by browsing to Anaconda2\lib\site-packages\ AMD confirming that there is only one zprocess folder there.

Also if you import zprocess in a python interaction session and print zprocess.__version__, that will help confirm you're using the right version as well.

Otherwise, an issue might be firewalling. Since a server is already running, it would be good if you could kill it, via Windows' task manager, and start one manually yourself. That way you can see whether any information is reaching the zlock server - because it will print it out on the terminal.

To unsubscribe from this group and stop receiving emails from it, send an email to labscriptsuite+unsubscribe@googlegroups.com.

[Pierre Colombo]

Thank you for your help : 

- zprocess is gone for good 

- import zprocess, zprocess.__version__ gives me 1.3.3 .

- i don't find the relevant process to kill the server do you have any idea of the name ?

I just noticed a stange thing :

- am I supposer to install all the suite on after the other by cloning the repo ? (https://bitbucket.org/labscript_suite/)

- or can i use the the installer ? https://bitbucket.org/labscript_suite/installer that what i did but it was updated only in2015 !

Best 

Pierre

[Chris Billington]

Windows task manager will likely just list the process as 'python', so if there is only one process named 'python', that will be it. Otherwise, there is some way (mouseover tooltip?) to view the process's command line arguments, which should be 'python -m zprocess.locking.locking'. Also, a reboot will get rid of it.

Nope, you are supposed to use the installer (by cloning its repository), as you did. It doesn't update often, because it is configured to install the latest version of each other component, rather than any specific version. So the labscript suite can change quite a bit without the installer changing at all.

To unsubscribe from this group and stop receiving emails from it, send an email to labscriptsuite+unsubscribe@googlegroups.com.

[Pierre Colombo]

Alright THANKS A LOT. I can start now : 

- lyse 

-runmanager

-runviewer 

I killed the server manually and did : python -m zprocess.locking and it is working :D ! 

Do you think I have to do the same trick every time I start ?

Now I have another error for blacs (before it didn't start ) : 

Traceback (most recent call last):

  File "C:\labscript_suite\blacs\__main__.py", line 687, in <module>

    connection_table = ConnectionTable(connection_table_h5_file)

  File "C:\labscript_suite\blacs\connections.py", line 27, in __init__

    with h5py.File(h5file,'r') as hdf5_file:

  File "C:\labscript_suite\labscript_utils\h5_lock.py", line 42, in __init__

    _orig_init(self, name, mode, driver, libver, **kwds)

  File "C:\Program Files\Anaconda2\lib\site-packages\h5py\_hl\files.py", line 272, in __init__

    fid = make_fid(name, mode, userblock_size, fapl, swmr=swmr)

  File "C:\Program Files\Anaconda2\lib\site-packages\h5py\_hl\files.py", line 92, in make_fid

    fid = h5f.open(name, flags, fapl=fapl)

  File "h5py\_objects.pyx", line 54, in h5py._objects.with_phil.wrapper (C:\Minonda\conda-bld\h5py_1474482483473\work\h5py\_objects.c:2705)

  File "h5py\_objects.pyx", line 55, in h5py._objects.with_phil.wrapper (C:\Minonda\conda-bld\h5py_1474482483473\work\h5py\_objects.c:2663)

  File "h5py\h5f.pyx", line 76, in h5py.h5f.open (C:\Minonda\conda-bld\h5py_1474482483473\work\h5py\h5f.c:1951)

IOError: Unable to open file (Unable to open file: name = 'c:\experiments\example_experiment\connectiontable.h5', errno = 2, error message = 'no such file or directory', flags = 0, o_flags = 0)

Do you have any ideas ?

[Pierre Colombo]

Am I supposed to create the connectiontable.h5 before launching blacs ?

[Philip Starkey]

Hi Pierre,

You do need to create a connection table h5 file for the first launch of BLACS. A good starting point is to compile the labscript/example.py file from runmanager, and copy the resulting h5 file to the location of your connection table as per the setting in the labconfig file. Further information on this can be found on our documentation page (the "Using BLACS" document is a good place to start), however please note that the labscript API documentation is somewhat out of date, particularly with the instantiation of PseudoclockDevices. The labscript/example.py file shows the updated way to instantiate PulseBlasters and PineBlasters and associated ClockLines. If you have any questions about this, please let us know!

Once you've made the connection table h5 file, and launched BLACS, you can recompile the connection table .py (as specified in your labconfig file) file from within BLACS, so you won't have to do the more complicated file copy every time you want to change the connection table.

Cheers,

Phil

--

Philip Starkey

School of Physics and Astronomy

Monash University

To unsubscribe from this group and stop receiving emails from it, send an email to labscriptsuite+unsubscribe@googlegroups.com.

[Pierre Colombo]

Hi Philip,

Thanks for your help, the point is i cannot compile example.py.

I get an error : 

[Pierre Colombo]

Hi,

If i try to generate via runmanger I get an server time out error. 

I have tryied to execute as an admin the runmanager and launch manually runmanger with admin privileges but I canno't get the hd5 file, i always get the server timed out error.

Thanks for your time

Best

Pierre

[Philip Starkey]

Hi Pierre,

If the timeout error is relating to sending the shot to BLACS, this is expected behaviour if you have "run shots" ticked in runmanager and BLACS is not running. Feel free to untick "run shots" in runmanager when you just want to produce the HDF5 file but not send it to BLACS for execution.

That said, regardless of the error in sending the files to BLACS, the HDF5 files are still being produced in the folder specified by "shot output folder" in runmanager.

Cheers,

Phil

--

Philip Starkey

School of Physics and Astronomy

Monash University

To unsubscribe from this group and stop receiving emails from it, send an email to labscriptsuite+unsubscribe@googlegroups.com.

[Chris Billington]

Hi Pierre,

Just about the zlock difficulties - no, you shouldn't have to launch the server yourself. So something is still not quite right, although the workaround of launching it yourself is obviously fixing things. So if everything else is fine it's ok to keep doing this, but I'm wary that other problems might still come up.

I notice that some of your error messages are coming from Python packages in C:\Program Files\Anaconda2, and others are coming from C:\Miniconda. Do you possible have multiple Python installations on this computer? If so, this might still leave room for multiple versions of packages being around (and might still explain the zprocess bug - perhaps zprocess 2.0 is in C:\Miniconda?). The labscript suite creates shortcuts and sets up paths based on whichever Python executable was used to run the installer, which, if you have since installed a different Python distribution, may not be the same as what you currently get by typing 'python' in a terminal. I admit I'm not totally sure how Anaconda and Miniconda interact - perhaps it's normal to have a Python distribution spread between the two folders, but I wouldn't have thought so.

So I just wanted to advise it can definitely cause issues of the type you were initially seeing to have multiple Pythons installed. To rectify, you can delete one of them, but then you may need to run the labscript suite installer again in order for Python to be able to find the labscript suite modules (the labscript suite installer adds its install directory to the Python search path). Running the installer again will not overwrite or delete any changes you have made to configuration files or anything in the <labscript_suite_install_folder>/userlib directory.

Cheers,

Chris

[Pierre Colombo]

Hi Chris, hi Philip,

Thanks for your help. I have configure the windows firewall to let the server run. Everything is working well.

Thanks again for your patience and help

Cheers, 

Pierre