3,035
edits
No edit summary |
mNo edit summary |
||
| Line 1: | Line 1: | ||
** Be aware of other patches in progress | ** Be aware of other patches in progress | ||
** Talk to each other to ensure that your patches won't cause regressions | ** Talk to each other to ensure that your patches won't cause regressions | ||
** Run the usual smoke tests and (ideall) some sort of automated script that can spot SNAFUs immedeately | ** Run the usual smoke tests and (ideall) some sort of automated script that can spot SNAFUs immedeately | ||
* Don't file IT requests for production updates every hour, and don't email them every 5 minutes (or text-message them either) -- your best bet is to group your updates and push them weekly. | * Don't file IT requests for production updates every hour, and don't email them every 5 minutes (or text-message them either) -- your best bet is to group your updates and push them weekly. | ||
= Tagging for production = | = Tagging for production = | ||
* Tag your last-known-working-copy as your ROLLBACK (if one doesn't already exist) | * Tag your last-known-working-copy as your ROLLBACK (if one doesn't already exist) | ||
<code> | |||
[morgamic@khan aus]$ cvs tag -f AUS2_ROLLBACK | [morgamic@khan aus]$ cvs tag -f AUS2_ROLLBACK | ||
</code> | |||
* Review LXR logs to make sure each update makes sense | * Review LXR logs to make sure each update makes sense | ||
* Apply your patches for the next batch | * Apply your patches for the next batch | ||
* When you're _SURE_ that this is the final batch of updates, you need to tag it as PRODUCTION, then check that baby in | * When you're _SURE_ that this is the final batch of updates, you need to tag it as PRODUCTION, then check that baby in | ||
<code> | |||
[morgamic@khan aus]$ cvs tag -F AUS2_PRODUCTION | [morgamic@khan aus]$ cvs tag -F AUS2_PRODUCTION | ||
cvs tag: Tagging . | cvs tag: Tagging . | ||
| Line 31: | Line 26: | ||
T inc/xml.class.php | T inc/xml.class.php | ||
[morgamic@khan aus]$ cvs ci | [morgamic@khan aus]$ cvs ci | ||
</code> | |||
= Pushing updates into production = | = Pushing updates into production = | ||
| Line 39: | Line 34: | ||
= Requesting an update in production = | = Requesting an update in production = | ||
* [https://bugzilla.mozilla.org/enter_bug.cgi?format=it_request&product=mozilla.org File an IT Request] when you are sure you are ready to go | * [https://bugzilla.mozilla.org/enter_bug.cgi?format=it_request&product=mozilla.org File an IT Request] when you are sure you are ready to go | ||
* Assign the appropriate severity -- note that critical or blocker pages the oncall. | * Assign the appropriate severity -- note that critical or blocker pages the oncall. | ||
= After the push = | = After the push = | ||
* Always be available after updates are pushed | * Always be available after updates are pushed | ||
* Sysadmins, if the developer did things right, you should be able to rollback by using: | * Sysadmins, if the developer did things right, you should be able to rollback by using: | ||
<code> | |||
[root@khan aus]$ cvs up -r AUS2_ROLLBACK | [root@khan aus]$ cvs up -r AUS2_ROLLBACK | ||
</code> | |||
edits