| | 1 | Power cycle a node, on "boss": |
| | 2 | {{{ |
| | 3 | power cycle <node> |
| | 4 | }}} |
| | 5 | |
| | 6 | Add (or remove) a user to (from) the wheel group: |
| | 7 | {{{ |
| | 8 | wap unixgroups -a [user] wheel |
| | 9 | Usage: /usr/testbed/sbin/unixgroups <-h | -p | < <-a | -r> uid gid...> > |
| | 10 | -h This message |
| | 11 | -p Print group information |
| | 12 | -a uid gid... Add a user to one (or more) groups |
| | 13 | -r uid gid... Remove a user from one (or more) groups |
| | 14 | }}} |
| | 15 | |
| | 16 | Check the syntax of an ns file: |
| | 17 | {{{ |
| | 18 | nscheck nsfile |
| | 19 | where: |
| | 20 | nsfile - Path to NS file you to wish check for parse errors |
| | 21 | }}} |
| | 22 | |
| | 23 | Mark nodes "hwdown", on "boss". |
| | 24 | {{{ |
| | 25 | (Make sure that /usr/testbed/sbin is in your path.) |
| | 26 | wap sched_reserve emulab-ops hwdown [list-of-nodes] |
| | 27 | ^ desired state |
| | 28 | |
| | 29 | nfree emulab-ops reloading [list-of-nodes] |
| | 30 | ^ current state |
| | 31 | }}} |
| | 32 | |
| | 33 | To allocate node(s) that are currently "free", on "boss": |
| | 34 | {{{ |
| | 35 | nalloc emulab-ops hwdown [list-of-nodes] |
| | 36 | ^ desired state |
| | 37 | |
| | 38 | Usage: nalloc [-v] [-p] <pid> <eid> <node> <node> <...> |
| | 39 | -p enables partial allocation mode |
| | 40 | -v enables debugging output |
| | 41 | }}} |
| | 42 | |
| | 43 | Get switch statistics (such as packet counts) for an experiment. On "users": |
| | 44 | {{{ |
| | 45 | portstats <proj> <exp> |
| | 46 | prints statistics for all experimental interfaces in an experiment. |
| | 47 | portstats -h |
| | 48 | lists the other options. |
| | 49 | }}} |
| | 50 | |
| | 51 | Load an image onto node(s) in an experiment. |
| | 52 | {{{ |
| | 53 | os_load [options] node [node ...] |
| | 54 | os_load [options] -e pid,eid |
| | 55 | where: |
| | 56 | -i - Specify image name; otherwise load default image |
| | 57 | -p - Specify project for finding image name (-i) |
| | 58 | -s - Do *not* wait for nodes to finish reloading |
| | 59 | -m - Specify internal image id (instead of -i and -p) |
| | 60 | -r - Do *not* reboot nodes; do that yourself |
| | 61 | -e - Reboot all nodes in an experiment |
| | 62 | -z 1 - Zero unused portions of the disk. |
| | 63 | (Currently only on "boss".) |
| | 64 | node - Node(s) to reboot (pcXXX) |
| | 65 | |
| | 66 | Wrapper Options: |
| | 67 | --help Display this help message |
| | 68 | --server Set the server hostname |
| | 69 | --login Set the login id (defaults to $USER) |
| | 70 | --debug Turn on semi-useful debugging |
| | 71 | }}} |
| | 72 | |
| | 73 | Schedule a Node's future assignment(s) |
| | 74 | {{{ |
| | 75 | wap sched_reserve <pid> <eid> <node> [<node> ...] |
| | 76 | }}} |
| | 77 | Schedule a OS installation |
| | 78 | {{{ |
| | 79 | wap sched_reload (see sched_reload --help) |
| | 80 | }}} |
| | 81 | |
| | 82 | Release a node from it's current assignment. |
| | 83 | {{{ |
| | 84 | wap nfree <pid> <eid> [<node>] [<node>...] |
| | 85 | ***don't forget the node name or you'll free |
| | 86 | ***all the nodes in the project/group |
| | 87 | }}} |
