Choose timezone
Your profile timezone:
Weekly Meeting
→
Europe/Zurich
Norton Room (RAL)
Norton Room
RAL
Description
<html>
<head>
</head>
<body>
- Past Events
- Catania
- Future Events
- 2nd EGEE User Forum - Manchester
- HPDC 2007 Workshop on Grid Monitoring
- Sixth UK e-Science All Hands Meeting - full paper: 16th April 2007.
- Actions
- Code Registry - Alastair - 27/4/2006
Registry is just about there - but without replication of course. Uses a dummy schema. Using Martin's remote registry service which needed some changes. Not yet described in design. It must also be protected from bad ACL files. Will remove context in Registry and use ThreadLocal. - Code Schema - Ming - 8/1/2007
Schema coding work has started Will use a single DB e.g. RGMA_Schema specified in the config file. The VDB tables has the VDB name prepended. Class framework and constructors done. The replication is now being worked on. - Refactor to use a single servlet for all R-GMA - Kush
-
11/12/2006
Code checked in and tests work. - Check that logging chapter of design is consistent with what has been recently requested by sysadmins - Steve - 1/1/2007
- Identify any of our components that are using the old
APIs. If
they are raise a bug and fix. - Antony - 8/1/2007
Started - found a problem with the flexi - which takes a named database. Can't make an instant change. - Ensure that the start call to Producer being added by
Alastair is
tested and the 2 new exceptions added by Ming are tested. -
Parminder - 8/1/2007
DONE for 1.5 and being migrated to HEAD API Java test is currently missing from HEAD. - Watch bug on job publishing tables with VO and
predicate - Antony
- 15/1/2007
22869 - Antony will agitate - Find why Alessandra's mon box is unhappy - Alastair - 15/1/2007 CERN monb001 is also not happy. Threads and memory etc run out. Runs out of memory in about 90 minutes.
- Set up a test to simulate 1000's of producers feeding
an archiver
- Parminder - 15/1/2007
Managed to provoke an out of memory exception on the server. Run with constant but tolerable load to see if it degrades. - Investigate Martin's assertion: "The current registry
also cannot
handle queries from the new code with table aliases - they are
converted into 'table AS alias' which it doesn't understand."
This needs checking in the new code and a possible fix for the
old. - Kush - 22/1/2007
Martin says: I *think* this was a problem I encountered with a query like
select t.col from table t
The SQL parser (the new one I think) was translating this into
select t.col from table AS t
which I believe is also valid SQL. However the old registry didn't understand this form and threw the query out as invalid. I think I flagged it up because it was a new consumer / old registry incompatibility.
Alastair and Kush to investigate further - Get UI info so that we can get back in the RTM - Antony - 1/2/2007 Antony to talk to Ales
- Talk to Philipa next week about maintenance of registry XML - Antony and Alastair - 1/2/2007
- Identify service details for new consumer to expose -
Antony -
1/2/2007
Kush has a list which he will circulate via e-mail Inspector will check version of compoments to ask the right questions. This is instead of modifying the existing producers as it appears to be safer. - Identify simple status info - Antony - 1/2/2007
- All 1.5 test changes need copying to HEAD. They should
first be
identified - Parminder - 9/2/2007
Need to identify changes in 1.5 since 1.5 was moved to HEAD - Need to look at the other APIs for parsing error (1.5 and HEAD) and provide tests (1.5 and HEAD) to ensure that the Java API bug is not present - ??? - 9/2/2007
- Improve release plan - Steve - 9/2/2007
- Add new new calls to list available VDBs to spec and design - Steve - 20/2/2007
- Add to the spec that both client and mon box names should be a stringified numeric IP addresses
- Update design for schema following e-mail - Steve - 20/2/2007
- Add memory management (JMX and MBeans) to HEAD - Kush - 20/2/2007
- Rule of thumb for how many mon boxes a site should run - Parminder - 20/2/2007
- Release Antony's OutOfMemory watching script - Antony - 20/2/2007
- Better algorithm for max HEAP size - check for user's ULIMITs - Antony - 20/2/2007
- Write test for client parser bug in C, C++ and Python
- Parminder
- 20/2/2007
The python work should be passed on if you are not happy with Python.
- Code Registry - Alastair - 27/4/2006
- Support list
- 12/3/2007 - Paventhan
- 19/3/2007 - Ming
- 26/3/2007 - Paventhan
- 2/4/2007 - Kush
- 9/4/2007 - Antony
- 16/4/2007 - Parminder
- Bugs
- Patches
- Release plan
- Plan for week ahead (was)
- Antony - RTM UI, service status proposal, scan of old tools, out of memory watching script and heap size setting, bug 23361
- Paventhan - Look into Antlr, RGMA and BDII with C++ get priority
- Alastair - Support, table aliases, make better use of ETICS, registry
- Ming - Schema, service status proposal
- Kush - Memory management, Remote calls and task q
- Parminder - List of changes in 1.5, add tests for client
parser bug, testing 1000 producers
- AOB
The agenda of this meeting is empty
