[EMAIL PROTECTED] wrote: > Hi, > > We have a device driver that uses OpenIPMI to interact with the BMC. > Our driver calls ipmi_request_settime() function to send platform > event messages to the BMC. However, we have noticed that the BMC did > not receive all the messages that were sent even when this command > returned success in each case.
Are you saying that the BMC did not queue all the events in the log, or are you sure it did not receive all the messages? There's a big difference. I'll assume that you are not sure about the messages, but know that the events are not in the event log. > > Interestingly, when we introduced a slight delay everytime before > calling this function, we saw that all the events were received by the > BMC. > > Could you help us figure out if we are calling this function in a > wrong way, or whether we are calling it too fast to be able to send > all events. Or is it that the message buffer used by OpenIPMI is not > large enough to hold all the messages, and therefore some messages get > dropped if the earlier messages were still there (not sent/processed). If you are getting valid (non-error) responses from the IPMI driver and the BMC, I'm pretty sure the messages are getting to the BMC and it is responding. That code is straightforward and well tested and I've used it to add a large number of the events to the event log from inside the kernel (panic situation). My first suspicion is that the BMC is not adding all the events to the event log even though it is responding that it did. -Corey ------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642 _______________________________________________ Openipmi-developer mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/openipmi-developer
