When VM naming number padding 1, machine “10” is named “vmname01” instead of “vmname10”

[gerikis]

When VM naming number padding 1, the first 9 machines are created correctly (vmname1-vmname9),
however machine “10” is named “vmname01” instead of “vmname10”.

Name must be machine1, machine2, machine3 ... machine10, machine11, and so on.

The padding is configured via " vm_name_suffix_counter_length" = 1 in the instance: RedHatConsulting_Utilities (v11.5) / Infrastructure / VM / Provisioning / Configuration / default.

The code used for naming is / RedHatConsulting_Utilities / Infrastructure / VM / Provisioning / Naming / vmname, on line 72:
"$n{#{suffix_counter_length}}"

"This has to do with how the $n{#} syntax works in CFME not specifically with this code. The miq-Utitlities code is just taking that vm_name_suffix_counter_length and putting it in $n{#{suffix_counter_length}} as seen here https://github.com/RedHatOfficial/miq-Utilities/blob/master/Automate/RedHatConsulting_Utilities/Infrastructure/VM/Provisioning/Naming.class/__methods__/vmname.rb#L72. From there the built in CFME naming functions that know how to read the $n(#) syntax take over. "

According to https://bugzilla.redhat.com/show_bug.cgi?id=1688672 this is working as designed in miq-Utilities .
I believe this must be fixed in RedHatConsulting_Utilities /

Another question if RH consulting script have another way how to generate a number after $n(1) is overflowing , or why RH consulting script rely on miq-utilities with generating a number.

https://github.com/RedHatOfficial/miq-Utilities/blob/master/Automate/RedHatConsulting_Utilities/Infrastructure/VM/Provisioning/Naming.class/__methods__/vmname.rb#L72

[itewk]

@gerikis this is not something the current team maintaining these scripts is prepared to take on. Solvign the problems of incremental parallelized naming without using $n{} is an insanely difficult and troublesome problem to solve and is already solved by $n{} and would be far simpler and make far more sense to update $n{} to increment more logically, even if it currently "works as designed".

I am willing to leave this open if this is something you want to try and make another alternative naming method that doesn't use the $n{} but if you are looking for the couple of volunteers who maintain this currently to take this on it's not likely to ever happen so I don't want anyone to get their hopes up.

[gerikis]

Thank you Ian, for the help ,highly appreciated . Hopefully later on somebody might takes a challenge (this issue) , or will taken consideration and will affect future developments.

[itewk]

@gerikis I suggest setting your padding to a succificently high number in the mean time.