 |
 |
Hi all, I am rewriting some monitoring and trying to correctly digest the XrdXrootdMonBuff. However, the documentation [1] for XROOTD_MON_READU/_READV is unclear to me. The docs suggest that one of the two is the "sum" of read operations for a readv, and the other is a sequence of individual read operations in a readv. However, *both* types are described as having multiple entires, being switched on by iov, and so on. The meaning of the flags implies READU is unpacked (i.e. a sequence of reads) whereas READV is not (i.e. entire read), but it is later mentioned that READV can also be followed by other READV of the same request. Is it safe to always construct a vector of READ entries for all adjacent READV/READU with the same info[].arg0.id[1], i.e. readv identifier? Cheers, Max [1] http://xrootd.org/doc/dev44/xrd_monitoring.htm#_Toc449037000 ######################################################################## Use REPLY-ALL to reply to list To unsubscribe from the XROOTD-L list, click the following link: https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-L&A=1