[TVB-782] HDF5 does not support concurrent read operations - Jira TVB

XML

Word

Printable

Details

Type: Bug
Status: Closed
Priority: Blocker
Resolution: Fixed
Affects Version/s: None
Fix Version/s: 1.0.1
Component/s: API (Simulator - Framework)
Labels:
None
Environment:
Fedora 16, Windows 7, MacOS

Description

Start large scale connectivity and select a surface for the 'Brain Surface' parameter. From time to time python will crash, and the OS logs should show a BAD_ACCESS exception.

Same bug reproduced when launching a burst with multiple portlets, and they were all reading from the same TimeSeries H5 file

Gliffy Diagrams

Attachments

Sub-Tasks

Create unit-test trying to reproduce our problem with HDF5

Closed

Bogdan Neacsa

100%

Evaluate how much value/delay using pyTable has.

Closed

Bogdan Neacsa

100%

Change hdf5 access python library from pytables to hdf5 in order to resolve the concurrency problems. Also do custom buffering mechanism in case of append_data since h5py performance is poor next to pytables.

Closed

Bogdan Neacsa

100%

Activity

People

Assignee:: Bogdan Neacsa

Reporter:: Bogdan Neacsa

Votes:: 0 Vote for this issue

Watchers:: 0 Start watching this issue

Dates

Created:: 15/Oct/12 5:21 PM

Updated:: 08/Nov/12 12:41 PM

Resolved:: 08/Nov/12 12:10 PM

Time Tracking

Estimated:

2w 1d

Remaining:

Logged:

2w 1h

Include sub-tasks