DECH: Some sites have concerns that failures in the central monitoring infrastructure negatively impact their availability figures. The SAM team have taken a number of steps to address the issue. In particular, sensors have been developed to monitor the health of the monitoring infrastructure itself, and the return codes of tests that fail to reach sites will be modified so as not to impact the availability figures.
DECH: Some sites are upgrading the resources from 32-bit architecture to 64-bit one. However, gLite does not work fully on the 64-bit architectures due to missing libraries, especially the software for WNs.
DECH: • DESY is concerned about how new VO requirements are handled; in particular those requirements which have an impact on the sites’ resources and on other VOs. It was proposed and accepted that every new requirement must be included in the VO’s CIC VO card and reported to the EGEE/WLCG operation meetings. It is not clear how the subsequent procedure then is. Prominent examples are the request for additional software installations on all sites which might affect other VOs and the request for scratch disk space per job of more than the 5GB agreed on earlier. Especially in the latter case sites might have problems to meet this requirement. Even more it is not clear how jobs which consume more space are treated to avoid blocking of Worker Nodes. In the past the VOs started to submit tickets to sites which do not fulfil new requirements rather than trying to find a common global solution with the operations group.
about the missing documentation for the administration of the WMS
concern about the missing documentation in the area of VO deregistration
KEK is working on the interoperation between gLite and NAREGI. The job adaptor prototype of NAREGI using SAGA will be released by the end of March 2009. In addiction, a new production system was designed and its subsystems (LCG/NARAEGI/iRODs) were deployed successfully. France is also doing work in this area. Are these efforts coordinated?
