Difference between revisions of "User:Natew"

From CCGB
Jump to: navigation, search
(Notes to Self)
Line 1: Line 1:
== Notes to Self ==
+
== Notes to Successor ==
  
* Look over cfengine, dns, dhcp configs
+
This is a note to whoever takes the role of sysadmin after me.
* Spec out hardware for wash/zoe
+
 
* Swathi's laptop
+
=== Gotchas ===
* Recompile AFS for SL 6.2
+
AFS reports incorrect/nonsensical values for the size of files, volumes, and partitions over 2TB.
* Look at backup script
+
This can be annoying e.g. when you try to move a volume to a server with gigabytes left and it turns out the volume is terabytes big.
* Put stuff into cfengine
+
Use {{code|volid}} to get the path of the volume on the AFS server and run {{code|du}} against it to get the correct volume size.
* Look into Ganglia unicast
+
 
* Saw this pop up while ssh'd into river:
+
There's a bunch of stuff you'll need admin access to, like
Oct  3 08:00:17 river.bx.psu.edu sendmail[791]: [ID 801593 mail.alert] q93C0FBv000791: Losing ./qfq93C0FBv000791: savemail panic
+
* user/admin Kerberos account
 +
* system:administrators AFS group
 +
* admin and bxadmins group in LDAP
 +
* probably other stuff I'm forgetting
  
 
== References ==
 
== References ==

Revision as of 12:14, 10 May 2013

Notes to Successor

This is a note to whoever takes the role of sysadmin after me.

Gotchas

AFS reports incorrect/nonsensical values for the size of files, volumes, and partitions over 2TB. This can be annoying e.g. when you try to move a volume to a server with gigabytes left and it turns out the volume is terabytes big. Use volid to get the path of the volume on the AFS server and run du against it to get the correct volume size.

There's a bunch of stuff you'll need admin access to, like

  • user/admin Kerberos account
  • system:administrators AFS group
  • admin and bxadmins group in LDAP
  • probably other stuff I'm forgetting

References