When the MDT is starting and in the WAITING_FOR_CLIENTS state the server is most likely processing the MGS config logs. This is done sequentially and involves setting up a connection to each of the MDT+OST servers. Normally this is fairly quick, however if there are down servers, or the connection otherwise fails, then the full timeout needs to be waited before proceeding on to setup the next connection. When there are a large number of unreachable servers this can take a very long time.
It would be helpful if some form of progress could be reported after the WAITING_FOR_CLIENTS line for each MDT. One option might be to add the number of records in the MGS mount log being process and the current record to the recovery_status file.
When the MDT is starting and in the WAITING_FOR_CLIENTS state the server is most likely processing the MGS config logs. This is done sequentially and involves setting up a connection to each of the MDT+OST servers. Normally this is fairly quick, however if there are down servers, or the connection otherwise fails, then the full timeout needs to be waited before proceeding on to setup the next connection. When there are a large number of unreachable servers this can take a very long time.
It would be helpful if some form of progress could be reported after the WAITING_FOR_CLIENTS line for each MDT. One option might be to add the number of records in the MGS mount log being process and the current record to the
recovery_statusfile.Example recovery_status file for MDT0000