Deploy: Upgrade script mapped drives

[John-Holt-Tessella]

If the upgrade script has mapped a drive and crashes it should not leave that drive behind. This includes when doing it through pushd.
Also check that mapping drives in this way does not disturb current mapped drives. As one of the server tests make sure that drives are correctly mapped (confirm with Chris):

1. Data drive D
2. Scripting drive U
3. One other?

Finally if running on an instrument that cannot access a network location (e.g. R78) it should give you an error message and an option of mapping using fed credentials (see deply script)

[ChrisM-S]

Ideally an instrument will only have D (data), U (user) and O (cOmmon) drives mapped, see here

[ChrisM-S]

These can be re-mapped using the domain version of the normal user account (details in the normal place). It is important that these are mapped manually from the windows explorer "Map Network Drive" menu on "Computer". Map each, one at a time giving path and username details. Ensure reconnect at logon is checked for each one too (if missed, disconnect and re-map with this). Doing a reboot and checking the drives are still mapped and have access to their shares is a good idea.

[ChrisM-S]

If possible, it might be an idea to only use UNC paths in the script i.e. \\machine\share\files, these can still be made accessible by a net use command but without a driver letter. e.g.
net use \\machine\share /user:user

[ChrisM-S]

oops - hit close button by mistake :-)