With messages like this: "No machine_name called 'OS_X_10.6.2_tracemonkey' can be found" For these branches: - tracemonkey - electrolysis - addonsmgr - places http://production-master02.build.mozilla.org:8010/builders/OS%20X%2010.6.2%20tracemonkey%20build http://production-master02.build.mozilla.org:8010/builders/OS%20X%2010.6.2%20electrolysis%20build http://production-master02.build.mozilla.org:8010/builders/OS%20X%2010.6.2%20addonsmgr%20build http://production-master.build.mozilla.org:8010/builders/OS%20X%2010.6.2%20places%20build?numbuilds=100

I'll work on the sql changes and then file a bug with IT to get them added to the graph server.

I did not see addons manager listed for any of the other OS's so left it off of the 10.6.2 set

Comment on attachment 449038 [details] [diff] [review] Add new tests for macosx64 to graph server sql addonsmgr is in the new master01: http://production-master01.build.mozilla.org:8010/builders/OS%20X%2010.6.2%20addonsmgr%20build/builds/0 Can you please add it? r=me with that Can you also attach an SQL paste to be run by catlee (for staging graph DB) and by IT (stage graph DB)? I have noticed (thanks to 0.8.0's new view) that this step consumes around 20 minutes for each build since it is waiting to time out. So this bug is a little more important than I thought at first.

adjusted patch to add the other OS items for AddOnsMgr

Comment on attachment 449064 [details] [diff] [review] Add AddOnsMgr items to sql statements perfect. I can land this for you once we have the sql statement landed on staging graph DB and the stage graph DB. NOTE, the sql statement should only be the 10.6 missing changes: insert into machines values (NULL,17,0,NULL,"OS_X_10.6.2_tracemonkey",1,unix_timestamp()); insert into machines values (NULL,17,0,NULL,"OS_X_10.6.2_tracemonkey_leak_test",1,unix_timestamp()); insert into machines values (NULL,17,0,NULL,"OS_X_10.6.2_places",1,unix_timestamp()); insert into machines values (NULL,17,0,NULL,"OS_X_10.6.2_places_leak_test",1,unix_timestamp()); insert into machines values (NULL,17,0,NULL,"OS_X_10.6.2_electrolysis",1,unix_timestamp()); insert into machines values (NULL,17,0,NULL,"OS_X_10.6.2_electrolysis_leak_test",1,unix_timestamp()); insert into machines values (NULL,17,0,NULL,"OS_X_10.6.2_addonsmgr",1,unix_timestamp()); insert into machines values (NULL,17,0,NULL,"OS_X_10.6.2_addonsmgr_leak_test",1,unix_timestamp()); Thanks bear.

macosx64 inserts only as text

Comment on attachment 449072 [details] 10.6.2 sql inserts only Looks good. Go ahead and ask catlee/alice to land this for you. After you see that the graph posts on staging work as expected you can file the IT bug for them to land it. Good job!

(In reply to comment #7) > (From update of attachment 449072 [details]) > Looks good. Go ahead and ask catlee/alice to land this for you. > After you see that the graph posts on staging work as expected you can file the > IT bug for them to land it. > > Good job! Ok, these inserts have been applied to graphs-stage

evidently this has already been applied to production - closing this bug

I've unhidden the build, debug, and nightly builders on TraceMonkey and Electrolysis. They are green except for debug which has the same hangs in trace-malloc as m-c has.