146 reputation
2
bio website kevinrae.com
location
age
visits member for 5 years, 7 months
seen Aug 23 '12 at 16:26

Jan
29
comment Are HP Nonstop servers becoming less common in the Enterprise?
We used several at my last shop. They migrated off the original Tandems onto the HP Nonstop hosts for support. The native applications had been running for over a decade on that platform and porting onto another system was just not cost-effective.
Nov
12
answered How to transfer a postgres database to a different server
Oct
9
answered Compare postgres db indexes between two databases
Aug
23
comment How to track superuser activities
I still think you should use a configuration management system. (puppet/cfengine/chef/systemimager/chef/etc...)
Aug
23
comment Best smartphone for sysadmins?
I like the G1. Better than an iPhone IMHO.
Aug
23
comment Best smartphone for sysadmins?
I love my G1. I need a real keyboard and I didn't like the virtual iPhone keyboard.
Aug
23
comment Managing a Linux cluster
pcampbell thanks for adding a Capistrano link. I didn't have enough points when I posted it.
Aug
23
comment Synchronizing processes accross two servers
You can roll your own workflow using Capistrano if you can write some basic ruby scripts.
Aug
23
answered ssh hangs without password prompt — works in root or other accounts
Aug
21
comment Automatic postgres backup
pg_dumpall is better than the other answers because this backs up your global database objects. (Ex: users, etc..) You will need these to completely recover your DB and if you have a large number of these than it can be non-trivial to recreate them. However, pg_dumpall could be overkill for you if you simply want to backup a single database instance within a multiple database cluster because pg_dumpall will backup every instance. So... as with most tech Q&A, it depends.
Aug
21
answered Managing the puppet filebucket?
Aug
21
awarded  Supporter
Jul
8
comment What is the best filesystem for insert performance on PostgreSQL?
+1 for Oskar. We just went from a server config where RAM was ~33% of the total size of the DB to a new machine where total RAM was greater than the size of the DB. Now we can cache the entire DB in memory. Our slowest SQL query is now 2 orders of magnitude faster.
Jul
8
answered LAMP-based “cloud” hosts like Mosso/Rackspace Cloud Sites and Cartika HAL?
Jun
19
answered Advice and tips for a junior sysadmin straight out of college?
Jun
18
answered Choosing a NAS system and things to watchout for
Jun
3
answered Tools for Troubleshooting SAN Performance Bottlenecks
Jun
2
answered Things every DBA for DB2 on Unix, Linux should know
Jun
2
answered Naming Conventions
May
27
awarded  Teacher