Hello. I'm trying to compile a new standard for our root disks, consistent with the EIS recommendations. I will be looking at a layout similar to this:
s0,/,16384 s1,swap,8192 s2,backup,- s3,,16384 s4,,, s5,,, s6,,60 s7,/var/opt,-free_hog- In the case of slice 6, that is used for metadb and slice 3 is the alternate root slice. So far it works fine, the only issue being that I am having trouble figuring out whether I can store sparse zones in /var/opt. Seems according to document 820-0178 that Solaris wants to make a copy of the zones to an alternate disk. The tricky part is I don't know beforehand whether the server will use zones or not. Ergo, I may have zones in /var/opt or I may use that for in-house written applications and executables. In other words, it is crucial that slice 7 stays shared between the boot environments but at the same time that they retain the ability to host zones. Again, I was unable to find a concrete example of this in 820-0178 and the attempts I've made to update the zones seems to suggest that some sort of local cache copy of each non-global zone is made by default. E.g. the /var/opt filesystem doesn't get mounted with slice 7 during construction of the alternate boot environment, but rather gets created in a directory off of slice 3 called var/opt. How can I get around this? Can you point me in the right direction, correct my ways or otherwise enlighten me as to what I am doing wrong? -Alex This message posted from opensolaris.org
