Summary list of ARs from this meeting: AR for Mark and Bryce: Transfer project info w/ supporting tests/scripts/docs. AR for Bryce (after he's caught up :-) upgrade the CPU test scripts to include more checking and write a "stress test" wrapper script AR for Joel: Post pointer to hardware management console downloads AR for Mary: check on firmware upgrade status AR for Natalie: Post a pointer to Li Shaohua's latest patch. AR for Martine: 1. Ask Silvester for an outline on the "on-demand computing" use case. 2. Ask for a list of possible subcases from the people that volunteer to help on the virtualization use case. 3. Send mail to platform management people to scope their interest in re-energizing their participation in the SIG. ################################# Attendees: Joel Schopp, IBM Mary Meredith, OSDL Mark Wong, OSDL Bryce Harrington, OSDL Bruce Vessey, Unisys Natalie Protasevich, Unisys Martine Silbermann, HP * Welcome to Bryce Harrington Bryce Harrington will be replacing Mark Wong and taking on the regression testing effort at OSDL. We thank Mark for his contributions to the SIG and welcome Bryce to the group. Mark will be spending some time w/ Bryce to transfer the project over. * Memory hotplug regression testing The system in the lab is finally booting but Mark didn't get a chance to run the regression test on it yet. It seems that the firmware that supports the hypervisor is not installed on that system so we're only able to create a single LPAR. Joel will send a pointer to the hardware management console bits needed on the partition to move the resources so that we can test the CPU hotplug code on this system while we're waiting to get the firmware upgraded. Bryce has submitted a request for use of an ia64 system in the lab. * CPU hotplug regression testing We had a discussion on how to check why a request for taking a CPU down was rejected. The scripts need to be updated to at least check that the values in sysfs are set to what we expect them to be. Also some sort of "wrapper script" needs to run the test script in a loop to provide for more stress testing. Natalie is waiting on Li Shaohua to stabilize his patch so she can merge her own changes. She will post a pointer to Li's latest patch so Bryce/Mark or anyone interested can download it and test it. * Use cases Silvester agreed to help on the use case for on-demand computing but he'll first put some effort on updating the usability section of the capabilities document. Martine will coordinate with him to avoid overlap with the virtualization use case. Martine will send mail to the people that volunteered to help on the virtualization use case to generate a list of possible subcases on that subject. * Paris meeting At the Paris meeting we'll give a brief status on the SIG and also get a chance to bring to the attention of the working groups specific items that we believe need more resources (hw, people, $$$, community attention, etc...). If there's any specific topic you'd like to be mentioned please send mail to Mary or Martine. We discussed the fact that the hotplug memory community is currently concentrating their efforts on getting patches accepted for hot-add and structural changes needed for hot-remove. Regression testing on hot-remove is still very desirable, in particular there are no known written tests for migration and allocation (testing mostly done "manually" by developers). The SIG could greatly help by generating such tests. The activity and participation level to the SIG from the platform management people has been very low lately. Martine will contact the interested parties to try to determine what/if/how we can re-energize that project. * Others The hotplug memory paper for OLS is written and is full of very useful information. It will not be accessible to the public until after OLS. Anyone who would like to see it earlier please contact Joel. Next meeting is scheduled for May 10th at 11:00am -12:00pm PST, 2:00pm - 3:00pm EST Thanks for your participation. Martine J. Silbermann
