User Tools

Site Tools


xythos:upgrade_times_log

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
xythos:upgrade_times_log [2011/08/01 09:45]
hmeij
xythos:upgrade_times_log [2011/08/01 11:46] (current)
hmeij [Qs]
Line 1: Line 1:
 ====== Qs ====== ====== Qs ======
-  can we get rid of the prompting? username/passwd, tables for datastores? + 
-  * what causes the internal id fix to take so long when content is present vs no content+Upgrade path:  Each time we took a copy of production, changed database name and user/password properties (xml properties file and inside database). Also staged content area (except in FIRST).  We have 5 questions: 
-  * + 
 +  #1 Can we get rid of the prompting when upgrading(username/passwd, tables for datastores ... just accept defaults provided) 
 + 
 + Summary of THIRD attempt: 
 + 
 +stage version 7.1.37.13\\ 
 +./fixServerPrincipleIDs -a [test|fix] ... server remains up !! test=7 mins, fix=20 hours \\ 
 +hotfix 7.1.37_sp2 (skipped)\\ 
 +upgrade to 7.2.88 (1:50)\\ 
 +hotfix 7.2.88.sp2 (skipped)\\ 
 +upgrade 7.2.91 (1:10)\\ 
 + 
 +  * #2 When executing fixServerPrincipleIDs script dev server can remain up (we tested). Is this true/advisable for production upgrade
 +  * #3 If yes above, should we while production is up, consider running in batches (directory trees) or all at once? 
 +  * #4 If answer is no, server must be down: when we apply the 7.2.88 upgrade (without fixServerPrincipleIDs script manually, this upgrade seems to automatically invoke it) script **and** upgrade take 19 hours (second attempt) versus script by itself only takes 20 hours (third attempt).  Why
 +  * #5 Can we skip the hotfix upgrades after running the script and go directly to 7.2.88, then to 7.2.91? 
 + 
 + 
 + 
 + 
  
 ====== UPGRADE TIMES BY MATT/HENK ====== ====== UPGRADE TIMES BY MATT/HENK ======
    
  
-FIRST: **copy of prod db on filer1, no content\\+**FIRST:** \\ 
 +**copy of prod db on filer1, no content\\
 exactly what hong did, achieved about similar times, roughly 6 hours + pre/post work**\\ exactly what hong did, achieved about similar times, roughly 6 hours + pre/post work**\\
 ...upgrade to 7.2.88...\\ ...upgrade to 7.2.88...\\
Line 26: Line 47:
 Jul 27, 2011 3:20:47 PM -- Exited successfully - 0:50 \\ Jul 27, 2011 3:20:47 PM -- Exited successfully - 0:50 \\
  
-SECOND: **copy of prod db on filer4, with content, roughly 20 hours + pre/post work**\\+**SECOND:**\\ 
 +**copy of prod db on filer4, with content, roughly 20 hours + pre/post work**\\
 ...upgrade to 7.2.88...\\ ...upgrade to 7.2.88...\\
 Jul 27, 2011 6:35:55 PM -- Xythos Database Schema Installer Version 7.2.88.1020 - (cumulative time)\\ Jul 27, 2011 6:35:55 PM -- Xythos Database Schema Installer Version 7.2.88.1020 - (cumulative time)\\
Line 45: Line 67:
  
  
-THIRD: **copy of prod db on filer4, with content **\\ +**THIRD:**\\ 
-...hotfix 7.1.37_24 on current version...\\ +**copy of prod db on filer4, with content **\\ 
-download and unpack +...hotfix 7.1.37_24 (contains the fixInternalPrincipalIDs script ) on current version...\\ 
-fix java path or stage old java +download and unpack\\ 
-./fixServerPrincipleIDs -a [test|fix] ... server remains uptest=7 mins, fix=20 hours \\ +fix java path or stage old java\\ 
-... 3:45 pm -> 11:45 am basically, 20 hours, NOTE this is longer than fixing it in step 2\\ +./fixServerPrincipleIDs -a [test|fix] ... server remains up !! test=7 mins, fix=20 hours \\ 
-... but then the server must be down, using the utility it appears it can remain up (verify with xythos)\\+...3:45 pm -> 11:45 am basically, 20 hours, NOTE this is longer than upgrade (includes fix) in step 2\\ 
 +...but then the server must be down, using the utility it appears it can remain up (verify with xythos)\\
 ...internal fix utility results: broken in home 210, broken in data 1,296 \\ ...internal fix utility results: broken in home 210, broken in data 1,296 \\
 +...upgrade to 7.1.37_sp2  (needed ???  time ???) \\
 ...upgrade to 7.2.88...\\ ...upgrade to 7.2.88...\\
-Rolled forward to 7.2.88 and it is predictably must faster when you run  +Jul 29, 2011 4:02:52 PM -- Xythos Database Schema Installer Version 7.2.88.1020 - (cumulative time)\\ 
-the installer. +Jul 29, 2011 4:35:55 PM -- Successfully upgraded the Xythos Global Schema - 0:35 \\ 
-...hotfix 7.2.88.sp2...\\ +Jul 29, 2011 5:00:20 PM -- Successfully upgraded the Xythos Document Store: 'Data Document Store - 1:00 \\ 
-upgraded 7.2.88 and then to 7.2.91 (with pauses  +Jul 29, 2011 5:28:44 PM -- Successfully upgraded the Xythos Document Store: 'Home Document Store - 1:30 \\ 
-inbetween), starting at 12:00PM and finishing at 3:00PM  +Jul 29, 2011 5:46:42 PM -- Workflow data upgrade completed - 1:45 \\ 
 +Jul 29, 2011 5:48:24 PM --     7_2_internal_id_fix - 1:50 \\ 
 +Jul 29, 2011 5:51:07 PM --     7_2_internal_id_fix - 1:50 \\ 
 +Jul 29, 2011 5:51:07 PM --     7_2_internal_id_fix - 1:50 \\ 
 +Jul 29, 2011 5:51:08 PM -- Exited successfully - 1:50 \\ 
 +...hotfix 7.2.88.sp2... (took 20 mins before, skip???)\\ 
 +...upgrade 7.2.91 (with pauses)\\ 
 +Jul 29, 2011 6:04:26 PM -- Xythos Database Schema Installer Version 7.2.91.1042 - (cumulative time) \\ 
 +Jul 29, 2011 6:05:11 PM -- Successfully upgraded the Xythos Global Schema - 1:00 \\ 
 +Jul 29, 2011 6:11:51 PM -- Successfully upgraded the Xythos Document Store: 'Data Document Store - 8:00 \\ 
 +Jul 29, 2011 6:51:44 PM -- Successfully upgraded the Xythos Document Store: 'Home Document Store - 45:00 \\ 
 +Jul 29, 2011 7:14:16 PM -- Exited successfully - 1:10 \\ 
 +...matt states: upgraded 7.2.88 and then to 7.2.91 (with pauses \\ 
 +...in between), starting at 12:00PM and finishing at 3:00PM -- check!)\\
  
  
xythos/upgrade_times_log.1312206355.txt.gz · Last modified: 2011/08/01 09:45 by hmeij