2
.\"___INFO__MARK_BEGIN__
4
.\" Copyright: 2004 by Sun Microsystems, Inc.
11
.\" Some handy macro definitions [from Tom Christensen's man(1) manual page].
13
.de SB \" small and bold
14
.if !"\\$1"" \\s-2\\fB\&\\$1\\s0\\fR\\$2 \\$3 \\$4 \\$5
17
.de T \" switch to typewriter font
18
.ft CW \" probably want CW if you don't have TA font
21
.de TY \" put $1 in typewriter font
30
.de M \" man page reference
31
\\fI\\$1\\fR\\|(\\$2)\\$3
33
.TH QHOST 1 "$Date$" "xxRELxx" "xxQS_NAMExx User Commands"
35
qhost \- show the status of xxQS_NAMExx hosts, queues, jobs
40
[ \fB\-F\fP [\fBresource_name,...\fP]
48
.B -l resource=val,...
56
shows the current status of the available xxQS_NAMExx hosts, queues and the
57
jobs associated with the queues. Selection options allow you
58
to get information about specific hosts, queues, jobs or users.
61
will display a list of all hosts without queue or job
67
.IP "\fB\-F\fP [ \fBresource_name,...\fP ]"
69
will present a detailed listing of the current
70
resource availability per host with respect to all resources (if the option
71
argument is omitted) or with respect to those resources contained in the
72
resource_name list. Please refer to the description of the
74
section \fBOUTPUT FORMATS\fP below for further detail.
77
Prints a listing of all options.
79
.IP "\fB\-h host_list\fP"
80
Prints a list of all hosts contained in host_list.
83
Prints all jobs running on the queues hosted by the shown hosts. This
84
switch calls \fB\-q\fP implicitly.
86
.IP "\fB\-l resource\fP[\fB=value\fP],..."
87
Defines the resources required by the hosts
88
on which information is requested.
89
Matching is performed on hosts.
93
Show information about the queues instances hosted by the displayed hosts.
95
.IP "\fB\-u user,...\fP"
96
Display information only on those jobs and queues
97
being associated with the users from the given user list.
100
This option can be used with all other options and changes the output to XML. The used
101
schemas are referenced in the XML output. The output is printed to stdout.
106
Depending on the presence or absence of the \fB\-q\fP or \fB\-F\fP and
107
\fB\-j\fP option three output formats need to be differentiated.
110
.SS "\fBDefault Format (without \-q, \-F and \-j)\fP"
111
For each host one line is printed. The output consists of
118
the Number of processors.
131
If the \fB\-q\fP option is supplied, each host status line also contains
132
extra lines for every queue hosted by the host consisting of,
136
the queue type \- one of B(atch), I(nteractive), C(heckpointing),
137
P(arallel), T(ransfer) or combinations thereof,
139
the number of used and available job slots,
141
the state of the queue \- one of
142
u(nknown) if the corresponding
143
.M xxqs_name_sxx_execd 8
144
cannot be contacted, a(larm), A(larm), C(alendar suspended), s(uspended),
145
S(ubordinate), d(isabled), D(isabled), E(rror) or
146
combinations thereof.
148
If the state is a(alarm) at least one of the load thresholds defined in the
149
\fIload_thresholds\fP list of the queue configuration (see
152
currently exceeded, which prevents from scheduling further jobs to that
155
As opposed to this, the state A(larm) indicates that at least one of the
156
suspend thresholds of the queue (see
158
is currently exceeded. This will result in jobs running in that queue being
159
successively suspended until no threshold is violated.
161
The states s(uspended) and d(isabled) can be assigned to queues and
164
command. Suspending a queue will cause all jobs executing in that queue to
167
The states D(isabled) and C(alendar suspended) indicate that the queue
168
has been disabled or suspended automatically via the calendar facility of
170
.M calendar_conf 5 ),
171
while the S(ubordinate) state
172
indicates, that the queue has been suspend via subordination to another
175
for details). When suspending a queue
176
(regardless of the cause) all jobs executing in that queue are suspended
179
If an E(rror) state is displayed for a queue,
180
.M xxqs_name_sxx_execd 8
181
on that host was unable to locate the
182
.M xxqs_name_sxx_shepherd 8
184
on that host in order to start a job. Please check the
185
error logfile of that
186
.M xxqs_name_sxx_execd 8
187
for leads on how to resolve the problem. Please enable the
188
queue afterwards via the \fB-c\fP option of the
192
If the \fB\-F\fP option was used, resource availability information is printed
193
following the host status line. For each resource (as selected in an option
194
argument to \fB\-F\fP or for all resources if the option argument was
195
omitted) a single line is displayed with the following format:
197
a one letter specifier indicating whether the current resource availability
198
value was dominated by either
200
`\fBg\fP' - a cluster global,
202
`\fBh\fP' - a host total or
204
a second one letter specifier indicating the source for the current resource
205
availability value, being one of
207
`\fBl\fP' - a load value reported for the
210
`\fBL\fP' - a load value for the resource after administrator
211
defined load scaling has been applied,
213
`\fBc\fP' - availability derived from
214
the consumable resources facility (see
218
availability definition derived from a non-consumable complex attribute or
219
a fixed resource limit.
221
after a colon the name of the resource on which information is displayed.
223
after an equal sign the current resource availability value.
225
The displayed availability values and the sources from which they derive are
226
always the minimum values of all possible combinations. Hence, for example,
227
a line of the form "qf:h_vmem=4G" indicates that a queue currently has a
228
maximum availability in virtual memory of 4 Gigabyte, where this value is a
229
fixed value (e.g. a resource limit in the queue configuration) and it is queue
230
dominated, i.e. the host in total may have more virtual memory available than
231
this, but the queue doesn't allow for more. Contrarily a line "hl:h_vmem=4G"
232
would also indicate an upper bound of 4 Gigabyte virtual memory
233
availability, but the limit would be derived from a load value currently
234
reported for the host. So while the queue might allow for jobs with higher
235
virtual memory requirements, the host on which this particular queue resides
236
currently only has 4 Gigabyte available.
238
After the queue status line (in case of \fB\-j\fP) a single line is printed
239
for each job running currently in this queue. Each job status
248
the status of the job \- one of t(ransfering),
249
r(unning), R(estarted), s(uspended), S(uspended) or T(hreshold) (see the
250
\fBReduced Format\fP section for detailed information),
252
the start date and time and the function of the job (MASTER
253
or SLAVE - only meaningful in case of a parallel job) and
255
the priority of the jobs.
258
.SH "ENVIRONMENTAL VARIABLES"
260
.IP "\fBxxQS_NAME_Sxx_ROOT\fP" 1.5i
261
Specifies the location of the xxQS_NAMExx standard configuration
264
.IP "\fBxxQS_NAME_Sxx_CELL\fP" 1.5i
265
If set, specifies the default xxQS_NAMExx cell. To address a xxQS_NAMExx
268
uses (in the order of precedence):
272
The name of the cell specified in the environment
273
variable xxQS_NAME_Sxx_CELL, if it is set.
275
The name of the default cell, i.e. \fBdefault\fP.
280
.IP "\fBxxQS_NAME_Sxx_DEBUG_LEVEL\fP" 1.5i
281
If set, specifies that debug information
282
should be written to stderr. In addition the level of
283
detail in which debug information is generated is defined.
285
.IP "\fBxxQS_NAME_Sxx_QMASTER_PORT\fP" 1.5i
286
If set, specifies the tcp port on which
287
.M xxqs_name_sxx_qmaster 8
288
is expected to listen for communication requests.
289
Most installations will use a services map entry for the
290
service "sge_qmaster" instead to define that port.
295
.ta \w'<xxqs_name_sxx_root>/ 'u
296
\fI<xxqs_name_sxx_root>/<cell>/common/act_qmaster\fP
297
xxQS_NAMExx master host file
302
.M xxqs_name_sxx_intro 1 ,
310
.M xxqs_name_sxx_execd 8 ,
311
.M xxqs_name_sxx_qmaster 8 ,
312
.M xxqs_name_sxx_shepherd 8 .
317
.M xxqs_name_sxx_intro 1
318
for a full statement of rights and permissions.