Skip to content
GitLab
Explore
Sign in
Primary navigation
Search or go to…
Project
S
Slurm
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Snippets
Build
Pipelines
Jobs
Pipeline schedules
Artifacts
Deploy
Releases
Package Registry
Model registry
Operate
Environments
Terraform modules
Monitor
Incidents
Service Desk
Analyze
Value stream analytics
Contributor analytics
CI/CD analytics
Repository analytics
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
Community forum
Contribute to GitLab
Provide feedback
Terms and privacy
Keyboard shortcuts
?
Snippets
Groups
Projects
Show more breadcrumbs
tud-zih-energy
Slurm
Commits
3dd6dc15
Commit
3dd6dc15
authored
21 years ago
by
Moe Jette
Browse files
Options
Downloads
Patches
Plain Diff
Reset priority of system held jobs (priority==1) when a non-responding
node begins to respond again.
parent
57050cab
No related branches found
No related tags found
No related merge requests found
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
src/slurmctld/node_mgr.c
+15
-7
15 additions, 7 deletions
src/slurmctld/node_mgr.c
with
15 additions
and
7 deletions
src/slurmctld/node_mgr.c
+
15
−
7
View file @
3dd6dc15
...
@@ -962,7 +962,8 @@ int update_node ( update_node_msg_t * update_node_msg )
...
@@ -962,7 +962,8 @@ int update_node ( update_node_msg_t * update_node_msg )
}
}
if
(
state_val
!=
NO_VAL
)
{
if
(
state_val
!=
NO_VAL
)
{
base_state
=
node_ptr
->
node_state
&
(
~
NODE_STATE_NO_RESPOND
);
base_state
=
node_ptr
->
node_state
&
(
~
NODE_STATE_NO_RESPOND
);
if
(
!
_valid_node_state_change
(
base_state
,
state_val
))
{
if
(
!
_valid_node_state_change
(
base_state
,
state_val
))
{
info
(
"Invalid node state transition requested "
info
(
"Invalid node state transition requested "
"for node %s from=%s to=%s"
,
"for node %s from=%s to=%s"
,
...
@@ -1000,7 +1001,8 @@ int update_node ( update_node_msg_t * update_node_msg )
...
@@ -1000,7 +1001,8 @@ int update_node ( update_node_msg_t * update_node_msg )
bit_clear
(
avail_node_bitmap
,
node_inx
);
bit_clear
(
avail_node_bitmap
,
node_inx
);
}
}
else
{
else
{
info
(
"Invalid node state specified %d"
,
state_val
);
info
(
"Invalid node state specified %d"
,
state_val
);
err_code
=
1
;
err_code
=
1
;
error_code
=
ESLURM_INVALID_NODE_STATE
;
error_code
=
ESLURM_INVALID_NODE_STATE
;
}
}
...
@@ -1010,7 +1012,8 @@ int update_node ( update_node_msg_t * update_node_msg )
...
@@ -1010,7 +1012,8 @@ int update_node ( update_node_msg_t * update_node_msg )
NODE_STATE_NO_RESPOND
;
NODE_STATE_NO_RESPOND
;
node_ptr
->
node_state
=
state_val
|
no_resp_flag
;
node_ptr
->
node_state
=
state_val
|
no_resp_flag
;
info
(
"update_node: node %s state set to %s"
,
info
(
"update_node: node %s state set to %s"
,
this_node_name
,
node_state_string
(
state_val
));
this_node_name
,
node_state_string
(
state_val
));
}
}
}
}
...
@@ -1121,7 +1124,10 @@ validate_node_specs (char *node_name, uint32_t cpus,
...
@@ -1121,7 +1124,10 @@ validate_node_specs (char *node_name, uint32_t cpus,
}
}
node_ptr
->
tmp_disk
=
tmp_disk
;
node_ptr
->
tmp_disk
=
tmp_disk
;
node_ptr
->
node_state
&=
(
uint16_t
)
(
~
NODE_STATE_NO_RESPOND
);
if
(
node_ptr
->
node_state
&
NODE_STATE_NO_RESPOND
)
{
reset_job_priority
();
node_ptr
->
node_state
&=
(
uint16_t
)
(
~
NODE_STATE_NO_RESPOND
);
}
if
(
error_code
)
{
if
(
error_code
)
{
if
((
node_ptr
->
node_state
!=
NODE_STATE_DRAINING
)
&&
if
((
node_ptr
->
node_state
!=
NODE_STATE_DRAINING
)
&&
(
node_ptr
->
node_state
!=
NODE_STATE_DRAINED
))
{
(
node_ptr
->
node_state
!=
NODE_STATE_DRAINED
))
{
...
@@ -1213,13 +1219,15 @@ void node_did_resp (char *name)
...
@@ -1213,13 +1219,15 @@ void node_did_resp (char *name)
last_node_update
=
time
(
NULL
);
last_node_update
=
time
(
NULL
);
node_ptr
->
last_response
=
time
(
NULL
);
node_ptr
->
last_response
=
time
(
NULL
);
resp_state
=
node_ptr
->
node_state
&
NODE_STATE_NO_RESPOND
;
resp_state
=
node_ptr
->
node_state
&
NODE_STATE_NO_RESPOND
;
node_ptr
->
node_state
&=
(
uint16_t
)
(
~
NODE_STATE_NO_RESPOND
);
if
(
resp_state
)
{
info
(
"Node %s now responding"
,
name
);
reset_job_priority
();
node_ptr
->
node_state
&=
(
uint16_t
)
(
~
NODE_STATE_NO_RESPOND
);
}
if
(
node_ptr
->
node_state
==
NODE_STATE_UNKNOWN
)
if
(
node_ptr
->
node_state
==
NODE_STATE_UNKNOWN
)
node_ptr
->
node_state
=
NODE_STATE_IDLE
;
node_ptr
->
node_state
=
NODE_STATE_IDLE
;
if
(
node_ptr
->
node_state
==
NODE_STATE_IDLE
)
if
(
node_ptr
->
node_state
==
NODE_STATE_IDLE
)
bit_set
(
idle_node_bitmap
,
node_inx
);
bit_set
(
idle_node_bitmap
,
node_inx
);
if
(
resp_state
)
info
(
"Node %s now responding"
,
name
);
if
((
node_ptr
->
node_state
==
NODE_STATE_DOWN
)
||
if
((
node_ptr
->
node_state
==
NODE_STATE_DOWN
)
||
(
node_ptr
->
node_state
==
NODE_STATE_DRAINING
)
||
(
node_ptr
->
node_state
==
NODE_STATE_DRAINING
)
||
(
node_ptr
->
node_state
==
NODE_STATE_DRAINED
))
(
node_ptr
->
node_state
==
NODE_STATE_DRAINED
))
...
...
This diff is collapsed.
Click to expand it.
Preview
0%
Loading
Try again
or
attach a new file
.
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Save comment
Cancel
Please
register
or
sign in
to comment