1 # EC2 configuration for Arvados Node Manager.
2 # All times are in seconds unless specified otherwise.
5 # The dispatcher can customize the start and stop procedure for
6 # cloud nodes. For example, the SLURM dispatcher drains nodes
7 # through SLURM before shutting them down.
10 # Node Manager will ensure that there are at least this many nodes
11 # running at all times.
14 # Node Manager will not start any compute nodes when at least this
18 # Poll EC2 nodes and Arvados for new information every N seconds.
21 # Polls have exponential backoff when services fail to respond.
22 # This is the longest time to wait between polls.
25 # If Node Manager can't succesfully poll a service for this long,
26 # it will never start or stop compute nodes, on the assumption that its
27 # information is too outdated.
28 poll_stale_after = 600
30 # "Node stale time" affects two related behaviors.
31 # 1. If a compute node has been running for at least this long, but it
32 # isn't paired with an Arvados node, do not shut it down, but leave it alone.
33 # This prevents the node manager from shutting down a node that might
34 # actually be doing work, but is having temporary trouble contacting the
36 # 2. When the Node Manager starts a new compute node, it will try to reuse
37 # an Arvados node that hasn't been updated for this long.
38 node_stale_after = 14400
40 # File path for Certificate Authorities
41 certs_file = /etc/ssl/certs/ca-certificates.crt
45 file = /var/log/arvados/node-manager.log
47 # Log level for most Node Manager messages.
48 # Choose one of DEBUG, INFO, WARNING, ERROR, or CRITICAL.
49 # WARNING lets you know when polling a service fails.
50 # INFO additionally lets you know when a compute node is started or stopped.
53 # You can also set different log levels for specific libraries.
54 # Pykka is the Node Manager's actor library.
55 # Setting this to DEBUG will display tracebacks for uncaught
56 # exceptions in the actors, but it's also very chatty.
59 # Setting apiclient to INFO will log the URL of every Arvados API request.
63 host = zyxwv.arvadosapi.com
67 # Accept an untrusted SSL certificate from the API server?
73 # It's usually most cost-effective to shut down compute nodes during narrow
74 # windows of time. For example, EC2 bills each node by the hour, so the best
75 # time to shut down a node is right before a new hour of uptime starts.
76 # Shutdown windows define these periods of time. These are windows in
77 # full minutes, separated by commas. Counting from the time the node is
78 # booted, the node WILL NOT shut down for N1 minutes; then it MAY shut down
79 # for N2 minutes; then it WILL NOT shut down for N3 minutes; and so on.
80 # For example, "54, 5, 1" means the node may shut down from the 54th to the
81 # 59th minute of each hour of uptime.
82 # Specify at least two windows. You can add as many as you need beyond that.
83 shutdown_windows = 54, 5, 1
92 # This section defines filters that find compute nodes.
93 # Tags that you specify here will automatically be added to nodes you create.
94 # Replace colons in Amazon filters with underscores
95 # (e.g., write "tag:mytag" as "tag_mytag").
96 instance-state-name = running
97 tag_arvados-class = dynamic-compute
101 # New compute nodes will send pings to Arvados at this host.
102 # You may specify a port, and use brackets to disambiguate IPv6 addresses.
103 ping_host = hostname:port
105 # Give the name of an SSH key on AWS...
108 # ... or a file path for an SSH key that can log in to the compute node.
109 # (One or the other, not both.)
112 # The EC2 IDs of the image and subnet compute nodes should use.
116 # Comma-separated EC2 IDs for the security group(s) assigned to each
118 security_groups = idstring1, idstring2
121 # You can define any number of Size sections to list EC2 sizes you're
122 # willing to use. The Node Manager should boot the cheapest size(s) that
123 # can run jobs in the queue (N.B.: defining more than one size has not been
125 # Each size section MUST define the number of cores it has. You may also
126 # want to define the number of mebibytes of scratch space for Crunch jobs.
127 # You can also override Amazon's provided data fields by setting the same