Quantcast
Channel: Comments for Clustering and High-Availability
Viewing all articles
Browse latest Browse all 797

re: Cluster Shared Volume (CSV) Inside Out

$
0
0

Vladimir,

Thank you, now it is clear! That is very valuable information.

So to get RW oplock, I need to break current RH that is held by MM/CC. The simplest way to do this - by writing. (It looks for me like general oplock problem: the only way to break R from other node is to make write operation)

But if I will write to the target file without having RW oplock, then what can stop Node3 to write at the same time(because it need to break R too) and to corrupt(by partial overwriting) data is such way?

The simplest workaround, I think, is just to read some existent byte from file and then overwrite it by itself.  This should break oplock and should not corrupt data in any case. I have made concurrent multinode testing - this works.  At the same time this looks weird. Is this a part of restrictions of oplock model? Maybe there are more elegant solution?


Viewing all articles
Browse latest Browse all 797

Trending Articles



<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>