I just submitted this as a bug to Parallels, but I thought I'd include it here for anyone with the same issue:
Upgrading from Plesk 10.3.1 to 11.0.9 via either /usr/local/psa/admin/bin/autoinstaller or via the panel, everything goes OK until the bootstrapper is run, and a database error occurs:
ERROR while trying to execute SQL query, the query was: REPLACE INTO ServiceNodes (uid=0(root) gid=0(root) groups=0(root), , ) VALUES (1, 'local', 'local');
The issue is line 2286 in /opt/psa/bootstrapper/pp11.0.9-bootstrapper/bootstrapper.sh, function db_upgrade_local_service_node_1012 - the line:
db_do_if "DESC ServiceNodes" "\$1 == \"name\"" "REPLACE INTO ServiceNodes (`id`, `name`, `transport`) VALUES (1, 'local', 'local');"
The backticks in that REPLACE INTO statement are being expanded in Debian, and I would suspect other distros as well, because the string is double-quoted. This causes the `id` to be replaced with the output of the 'id' command and create an invalid SQL statement. If the backticks are removed or escaped (\`), and I rerun bootstrapper.sh with the repair option, it proceeds and finishes.
This has happened on four servers, so it's consistent across all of my installs. Strangely, when I originally tested the upgrade, this worked fine - but I suspect that it was an earlier version of 11.
I suspect that I'm not the only person with this issue, so I thought I'd report.
Upgrading from Plesk 10.3.1 to 11.0.9 via either /usr/local/psa/admin/bin/autoinstaller or via the panel, everything goes OK until the bootstrapper is run, and a database error occurs:
ERROR while trying to execute SQL query, the query was: REPLACE INTO ServiceNodes (uid=0(root) gid=0(root) groups=0(root), , ) VALUES (1, 'local', 'local');
The issue is line 2286 in /opt/psa/bootstrapper/pp11.0.9-bootstrapper/bootstrapper.sh, function db_upgrade_local_service_node_1012 - the line:
db_do_if "DESC ServiceNodes" "\$1 == \"name\"" "REPLACE INTO ServiceNodes (`id`, `name`, `transport`) VALUES (1, 'local', 'local');"
The backticks in that REPLACE INTO statement are being expanded in Debian, and I would suspect other distros as well, because the string is double-quoted. This causes the `id` to be replaced with the output of the 'id' command and create an invalid SQL statement. If the backticks are removed or escaped (\`), and I rerun bootstrapper.sh with the repair option, it proceeds and finishes.
This has happened on four servers, so it's consistent across all of my installs. Strangely, when I originally tested the upgrade, this worked fine - but I suspect that it was an earlier version of 11.
I suspect that I'm not the only person with this issue, so I thought I'd report.