Sixth UK e-Science All Hands Meeting.
Antony will attend and Steve may have a day there
EGEE07 Budapest
Antony and Steve to attend
Actions
Alastair
Ensure 3.1 is correctly specified
Juha has changed configuration but it has raised a
question about how to set CLASSPATH for servictool, gin etc which rely
upon a symlink which is no longer being generated. Need either a
symlink or some other solution. Also include new tomcat checking
component. addressing bug 24643. Still no definitive
answer. Alastair will follow up
Building and testing new code
Registry tests now pass - but need to make small changes to follow
evolving spec. Adjusting unit tests for the registry to avoid
dependency on schema.. No need to test the
registry's non-forwarding
functions Will then move onto
consumer and schema. Will also need to modify configuration scripts.
Investigate the configuration parameters for MySQL and
for
Tomcat
Tried on RGMA17 but it does not work because the
parameters are for the wrong JVM
RGMA_HOME value
is built into the RPM. Should move to CATALINA_OPTS
Antony
Check with Matt Hodges on what he is doing with
Nagios and rgma-server-check
No reply yet
Check code and design of task Q
In progress. One algorithm is not quite right and error handling needs
improving - most of design matches the code.
Give info to Stephen Burke for status/version etc
- New
#27808
GIN not handling GlueServiceAssociation properly
#22205
Upgrade
to Glue schema 1.3
Should address Paventhan's bug and add in the meta
data columns at the same time. We must first be very clear about the
types of the columns.
Add metadata columns to all tables in
consultation with users
Patch
#1086 is in production so will need to tell people
to start
using the new exception soon
Get UI info so that we can get back in the RTM
Now we must wait for 3.1 to be deployed. Need to look at the
C++ header file on a WN to find
out. In future each API component should deposit a file with the
version number.
Arun
Read spec, design and bits of code - with emphasis on
primary producer
In progress
Produce sequence diagrams for PP
Declare table and done a bit of coding
Check the tuples store code for consistency with the
design
and spec
TupleStoreManager should be a singleton and there are some
discrepancies between design and code
Plan migration from physical DB names to logical
- Transferred from Antony
Ming
Write schema system tests
Now installed on 3 machines - most basic
operations work. Ming will concentrate on the tests on a single machine
first so that other components can be tested. Some tests pass - but not
all. There is some uncertainty in how "not null" columns were being
handled and should be handled. Authorization rules - need some
attention when they are not present. VDB Name - if not set
change
the servlet code to create a default VDBName - and raise a bug against
the python API.
#19645 The
Sub exception org.edg.info.ReadTimedOutException is an inappropriate
type
Parminder
Convert tests to python
Almost
done the non-parallel tests. Just checking that everything
agrees with original bash scripts.
See if we can use SLC3 machines - then can phase out .sh
files.
Steve
Logging document with Ales
Accepted by Ales - now with Romain. Still under
discussion. Still there are problems - no clear progress. Now
stalled.
Dependency challenge work
Reports on R-GMA and SD sent. Report on DGAS received. More
reports received
CHEP07 paper
Test that new streaming really works - given that we have
no new producers
It is here as a placeholder
Track patch #1093
for GIN
Now in certification (3.0.3 RC1)
Track patch #1212
Optimisations to the R-GMA registry as a fix for bug #27510
3.0.2 PPS Update 37
Track patch #1211 R-GMA
CLI fix for Bug #27504 Removed getAllTable call from __init__
3.0.2 PPS Update 37
Add to the spec that both client and mon box names
should be a stringified numeric IP addresses
Understand ping
Decide
on what to do with exceptions (2 or many and what about the types?).
Try to make what is says in the spec consistent with reality
Check for consistency between design and code
for
consumer