| | 33 | # |
|---|
| | 34 | # |
|---|
| | 35 | # HINT: To be very sure that your DRb server and application are always using |
|---|
| | 36 | # the same model and schema versions, and you never lose any index updates because |
|---|
| | 37 | # of the DRb server being restarted in that moment, use the following sequence |
|---|
| | 38 | # to update your application: |
|---|
| | 39 | # |
|---|
| | 40 | # cap deploy:stop deploy:update deploy:migrate deploy:start |
|---|
| | 41 | # |
|---|
| | 42 | # That will stop the DRb server after stopping your application, and bring it |
|---|
| | 43 | # up before starting the application again. Plus they'll never use different |
|---|
| | 44 | # versions of model classes (which might happen otherwise) |
|---|
| | 45 | # Downside: Your downtime is a bit longer than with the usual deploy, so be sure to |
|---|
| | 46 | # put up some maintenance page for the meantime. |
|---|
| 81 | | after "deploy:stop", "ferret:stop" |
|---|
| 82 | | after "deploy:start", "ferret:start" |
|---|
| 83 | | after "deploy:restart", "ferret:restart" |
|---|
| | 95 | after "deploy:stop", "ferret:stop" |
|---|
| | 96 | before "deploy:start", "ferret:start" |
|---|
| | 97 | |
|---|
| | 98 | before "deploy:restart", "ferret:stop" |
|---|
| | 99 | after "deploy:restart", "ferret:start" |
|---|