Facebook
Twitter
burnley
Regular Pleskian
Hi,
Just finished a Plesk Linux -> Plesk Linux migration of a fairly large server using version 2.7.9 and I must confess I'm pretty impressed with the improvements, especially on the reliability front. And the fact that it's highly scriptable and written in Python makes my job much easier to tailor it to our environment. Based on the most recent experience, here are few suggestions.
Feature required: Hooks configurable per client, per domain, including subdomains and domain aliases: pre, post. Exported variables:
- Stage: PRECLIENTMIGRATE, POSTCLIENTMIGRATE, PREDOMAINMIGRATE, POSTDOMAINMIGRATE, PREDOMALIASMIGRATE, POSTDOMALIASMIGRATE, etc...
- Plesk client name: PLESKCLNAME, aka "login" field in psa.clients
- Plesk subscriptions & domains, subdomains, domain aliases.
These variables can be returned as Python dictionary / json etc.
Data consistency improvements:
- Use "--delete" rsync switch when copying emails. This switch *MUST* be used to avoid the mail duplicates.
Speed improvements:
- Ability to disable "chown -R popuser
opuser" via an option. Useful when both source & target servers are using the same uid:gid for popuser.
- Ability to disable disk space checks. Using --skip-main-node-disk-space-checks for "transfer-accounts" doesn't seem to work, or I don't know how to use it.
- Ability to disable platform dependent code. E.g., if both source & target servers are running Plesk on Linux, I'm not expecting the migrator to run Windows specific code.
None of the 3 operations above is needed in our environment when doing Linux -> Linux migration and the speed improvement for large batches can be quite significant.
What do you think?
Just finished a Plesk Linux -> Plesk Linux migration of a fairly large server using version 2.7.9 and I must confess I'm pretty impressed with the improvements, especially on the reliability front. And the fact that it's highly scriptable and written in Python makes my job much easier to tailor it to our environment. Based on the most recent experience, here are few suggestions.
Feature required: Hooks configurable per client, per domain, including subdomains and domain aliases: pre, post. Exported variables:
- Stage: PRECLIENTMIGRATE, POSTCLIENTMIGRATE, PREDOMAINMIGRATE, POSTDOMAINMIGRATE, PREDOMALIASMIGRATE, POSTDOMALIASMIGRATE, etc...
- Plesk client name: PLESKCLNAME, aka "login" field in psa.clients
- Plesk subscriptions & domains, subdomains, domain aliases.
These variables can be returned as Python dictionary / json etc.
Data consistency improvements:
- Use "--delete" rsync switch when copying emails. This switch *MUST* be used to avoid the mail duplicates.
Speed improvements:
- Ability to disable "chown -R popuser
opuser" via an option. Useful when both source & target servers are using the same uid:gid for popuser.- Ability to disable disk space checks. Using --skip-main-node-disk-space-checks for "transfer-accounts" doesn't seem to work, or I don't know how to use it.
- Ability to disable platform dependent code. E.g., if both source & target servers are running Plesk on Linux, I'm not expecting the migrator to run Windows specific code.
None of the 3 operations above is needed in our environment when doing Linux -> Linux migration and the speed improvement for large batches can be quite significant.
What do you think?
