09.03.2015 Views

VSAN-Troubleshooting-Reference-Manual

VSAN-Troubleshooting-Reference-Manual

VSAN-Troubleshooting-Reference-Manual

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Diagnostics and <strong>Troubleshooting</strong> <strong>Reference</strong> <strong>Manual</strong> – Virtual SAN<br />

The block size that Virtual SAN uses to write to the flash cache device is 4KB. The<br />

block size that Virtual SAN uses when de-staging to magnetic disk is 1MB, which is<br />

also the stripe size used on magnetic disks and flash devices when they are used for<br />

capacity in version 6.0.<br />

The Virtual SAN flash is split into a read buffer and a write cache, 70% read and<br />

30% write. One common question that gets asked about the anatomy of a write<br />

operation is whether or not there is a subsequent write to the read cache so we<br />

don't need to re-read the currently written data from the magnetic disk.<br />

In Virtual SAN 5.5, the data in the read cache that may be affected by a new write is<br />

invalidated, and if the VM issues a read for that data block before it has been<br />

destaged to magnetic disk, we will read that data from the write buffer. A<br />

subsequent read after destaging may then incur a read cache miss.<br />

In Virtual SAN 6.0, this behavior has been changed. When the data is to be destaged<br />

to magnetic disk from the write buffer, the corresponding read cache entry (if it<br />

exists) is updated with the most recent data. Cache entries are not invalidated<br />

unless the cache decides that they're no longer useful. This should mean less<br />

magnetic disk I/O in Virtual SAN 6.0 when compared to Virtual SAN 5.5.<br />

V M W A R E S T O R A G E B U D O C U M E N T A T I O N / 2 4 6

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!