c372c8405e
vmmd: dump state (unikernel configurations) on create+destroy, restore from state on startup
2019-01-20 22:17:59 +01:00
1a288d2937
vmm_vmmd: pass success reply to handle_create fully applied
2019-01-20 22:09:33 +01:00
4d3bb777e1
vmm_vmmd: introduce and use type 'a create
2019-01-20 22:09:33 +01:00
e58ab236b7
vmmd: lock self_destruct with a mutex, respect result of killall (nothing killed, nothing to do)
2019-01-20 22:09:33 +01:00
d4e31da27f
vmmd: setup and teardown stat by create continuation, and vmm_vmmd.handle_shutdown
2019-01-20 22:09:33 +01:00
fdcea94a0b
vmmd: store waiter, not tasks in Vmm_vmmd.t -- create task and waiter on demand (destroy / create --force), instead of at each unikernel creation
2019-01-20 22:09:33 +01:00
fc63a89429
Vmm_unix: use Sys.sigterm (and rely on caml_convert_signal_number to be cross-platform)
2019-01-20 20:37:30 +01:00
dfd22be62b
Vmm_unix: don't reset image of a config
2019-01-20 20:35:55 +01:00
accb21b8b7
Vmm_unix.exec: rename vm to config
2019-01-20 20:35:26 +01:00
d30e118c76
Vmm_core.Name: use 'vm:' instead of 'name' as literal string to print
2019-01-20 20:31:40 +01:00
15e1f5ecf6
vmm_unix: ensure lazy uname being called only once
2019-01-20 20:28:51 +01:00
7b8f8fafbd
close fd early, no need to carry file descriptors around
2019-01-20 17:43:44 +01:00
094922f6b0
vmmd: teardown gracefully (kill all vms), install SIGTERM handler to teardown
2019-01-18 01:14:11 +01:00
58bd77bc5f
stats: pass bridge device through, vmm device name as well to allow arbitrary bhyve statistics, vmmc_local: add stats_add and stats_remove subcommands
2019-01-15 00:25:59 +01:00
fd4a5a5e22
inspect kinfo_proc structure for information about:
...
- virtual size
- resident size (in pages)
- text size (in pages)
- data size (in pages)
- stack size (in pages)
develop independent vmmc_stat for testing (not installed)
2019-01-06 01:12:56 +01:00
030f5aa379
vmm_unix, create_block: first create file, and then truncate
2018-12-07 00:07:16 +01:00
bda342f136
re-support parsing of AV2 log entries, document path forward strategy
2018-12-06 22:55:13 +01:00
e28ea84548
Log: new event `Hup
2018-12-06 22:53:15 +01:00
b5a068555c
vmmd_tlS_inetd
2018-11-23 21:07:36 +01:00
c8f1030403
rename Vm to Unikernel
2018-11-13 01:02:05 +01:00
85372b0c7e
rework resources: now block, vms, and policies are in separate tries
2018-11-13 00:06:43 +01:00
b5c9cdea6a
cleanups
2018-11-12 22:19:39 +01:00
dec32e6247
bump wire version
2018-11-12 22:11:06 +01:00
8ccda0e410
refactor bridge: use a string instead of a complicated thing
2018-11-12 22:07:45 +01:00
2e7f2730a2
move Vm to submodule
2018-11-11 03:24:50 +01:00
561ba5c5df
put Policy in a submodule
2018-11-11 03:09:37 +01:00
89a1d30154
cleanups in respect to directories and scope
2018-11-11 02:33:00 +01:00
43379d6d9d
rename Vmm_core.id to Vmm_core.Name.t and make it private - also check constructors to fit into 20 chars ldh (and in Vmm_tls max depth = 10)
2018-11-11 01:44:31 +01:00
6dcde8eb68
block device support
2018-11-11 00:01:56 +01:00
6945d21422
Vmm_core.drop_super / is_sub_id: fix function (used to assume reverse ordered labels)
2018-11-09 01:27:22 +01:00
75372a792f
fix resource policies. it was checking too many vms:
...
vm foo.bar is active with 32mb
add_policy bar --mem 16 <- failed :/
what is checked on add_policy <id> <new-policy>?
- all policies above <id> that <new policy> is a sub-policy
- all policies below <id> that each is a sub-policy of <new-policy>
- resource usage of vms below <id> is within <new-policy> limits (number of vms, memory, network access, cpuids)
2018-11-03 00:05:10 +01:00
79068c8abf
error logging on error in handle_command
2018-11-03 00:04:47 +01:00
0c58ebeedf
lower debug level for end of file
2018-11-01 01:54:10 +01:00
9f674f7e6f
forgot loop on this exit in add_policy
2018-11-01 01:23:45 +01:00
7c34c61d43
vmmd_tls needs looping behaviour after adding a policy to start vm
2018-10-31 23:03:30 +01:00
c669be8e02
address most of @cfcs comments
2018-10-29 17:14:51 +01:00
a124b3eb30
upgrade to recent decompress
2018-10-29 00:05:55 +01:00
2b85c65dd8
minor fixes from testing: do not require vm to be present for force-create, fix id generation in vmm_tls, use 32mb memory for unikernels by default
2018-10-28 23:06:15 +01:00
9191d2cf9a
drop version AV0, AV1; refactor vmm_asn
2018-10-28 22:52:20 +01:00
0f9375dc29
use oid 42 again
2018-10-28 22:39:31 +01:00
947b82f4f0
vmm_tls: ensure that add_policy commands carry a non-empty name
2018-10-28 22:30:08 +01:00
34291dbe65
vmmp_request
2018-10-28 22:14:39 +01:00
40519afbb7
issue policy_add commands by vmmd_tls for certificate chain
2018-10-28 20:50:10 +01:00
8ab37d6b3b
resources: remove_vm and remove_policy - no need to intertwine into a single remove
2018-10-28 19:50:48 +01:00
7b8f2cf802
add policy does nothing when received policy is equal to stored one
2018-10-28 19:41:06 +01:00
296b7a9b01
vmmd_tls: close sockets appropriately
2018-10-28 19:19:38 +01:00
5e921d7345
skip empty common names in vmm_tls
2018-10-28 19:04:24 +01:00
8f02d8263d
wip: vmmc_bistro
2018-10-28 02:03:27 +02:00
a60f866f70
fewer lists, read replies (to sockets) in vmmd
2018-10-26 21:30:54 +02:00
01f933702d
move stuff around
2018-10-26 21:30:54 +02:00
85a507db54
whitelist commands accepted via tls certificate
2018-10-26 21:29:59 +02:00
992e1b0a2b
- Vmm_ring is now polymorph (alows to store log_entry :D)
...
- Vmm_console/log/stats do not read multiple times
console_add loops
console_subscribe terminates (a stream of messages is sent)
log data stream loops
log_subscribe terminates (a stream of data is sent)
stat_add loops
stat_remove loops
stat_subscribe terminates (a stream of stats is sent)
terminates means: reads once more, and closes socket after second read returned
loop processes further incoming data
2018-10-26 21:29:59 +02:00
b55281d1e5
include version in log_entries on disk, read log file on startup (and write events to ring store)
2018-10-26 21:29:59 +02:00
cdae37b0bf
interface for vmm_ring
2018-10-26 21:29:59 +02:00
04367421bf
since argument for log_subscribe and console_subscribe
2018-10-26 21:29:59 +02:00
89fea934a7
copyright
2018-10-26 21:29:59 +02:00
a064c7f58e
move more stuff around
2018-10-26 21:29:59 +02:00
d513269453
move stuff into vmm_commands
2018-10-26 21:29:59 +02:00
6f18f1bfff
type data for streamed thingies
2018-10-26 21:29:59 +02:00
ce0c42fa77
more cleanups
2018-10-26 21:29:59 +02:00
46548418cd
minor cleanup: stats type
2018-10-26 21:29:59 +02:00
f3c67f626a
more cleanups
2018-10-26 21:29:59 +02:00
d896d89bba
.
2018-10-26 21:29:59 +02:00
f5ce2d8826
reuse commands from Vmm_asn.wire_commands for certificates
2018-10-26 21:29:59 +02:00
d6c87bacde
minor tweaks
2018-10-26 21:29:59 +02:00
a08f35ee5e
cleanups
2018-10-26 21:29:59 +02:00
811f3abc50
adjustments
2018-10-26 21:29:59 +02:00
c399501a18
get rid of vm_config.vname
2018-10-26 21:29:59 +02:00
0441b8ab25
tls endpoint
2018-10-26 21:29:59 +02:00
f939ff5a58
influx stats
2018-10-26 21:29:59 +02:00
1d4d7509dc
remove vmm_wire, use asn.1
2018-10-26 21:29:59 +02:00
51a0344477
fix warnings
2018-10-26 21:29:59 +02:00
2239aafdb7
revive vmm_client
2018-10-26 21:29:59 +02:00
bcb280aa00
refactor commands into vmm_commands
2018-10-26 21:29:59 +02:00
efc043cd5c
fix 'vmmc info'
2018-10-26 21:29:59 +02:00
182e2ae10c
policies:
...
vmmc now has more subcommands
- policy [-n name] returns all policies in name and below
- add_policy [-n name] [--cpu cpuid] [--mem mem] [--bridge bridge] [--block size] adds a policy
- remove [-n name] removes policy at name
policy is just the same which is in vmm_req_delegation, and vmm_resources now check them:
- you cannot insert a subpolicy violating the prefix
- you cannot insert a policy which would forbid current resource usage
- you cannot insert a policy with which any subpolicy would be invalid
- you can adjust (increase/decrease) a policy if the above invariants are kept
implement "force create" directly in vmmd: much nicer to
- check resource constraints,
- kill vm potentially,
- and create a new vm,
all as single transaction.
2018-10-26 21:29:59 +02:00
ea83013068
delegation -> policy
2018-10-26 21:29:59 +02:00
e413b8c99a
remove naming struggle in vm_config and Log.hdr
2018-10-26 21:29:59 +02:00
4c5a795a3b
console fifos in separate directory
2018-10-07 01:24:33 +00:00
133884faf4
log, stats and console socket go in their own directory
2018-10-07 00:32:25 +00:00
38094a53e3
use vmm_trie in log and stat, cleanups
2018-09-28 22:44:38 +02:00
02f8d94db8
s/ukvm/hvt/
2018-09-21 22:31:04 +02:00
38b98ab318
minor
2018-09-20 23:19:55 +02:00
99ba1c5e4b
stats are back now! no longer two pullers, but now with one pusher :)
2018-09-20 22:53:42 +02:00
e7b4742964
less is more, also unify default socket paths
...
and vmmc console command
2018-09-19 21:53:18 +02:00
bd10209297
wip, vmmc and vmmd talk with each other!
2018-09-09 20:52:04 +02:00
9ec69e23cc
rename Vmm_commands to Vmm_unix
2018-07-07 23:14:49 +02:00
c9afb6f5dc
CLOEXEC for all the openfile() calls to avoid leaking the fifo's (created and opened by vmmd itself) to ukvm-bin started later
2018-04-30 00:20:28 +02:00
2bf1b96178
get rid of FreeBSD kludge by using devfs!
2018-04-26 22:17:54 +02:00
0583fbfaf1
stats: transmit vmid in add/remove/stats, pid only in add
...
don't use /tmp anymore, but /var/run/albatross for fifos + sockets + vm images,
and /var/db/albatross for ukvm-bin and crls, and /var/log/albatross for logging
vmm_console/vmm_log/vmm_stats_lwt: delete socket on startup if it exists
vmm_influxdb_stats: connects to vmm_stats socket and pushes every interval in
influxdb line format via tcp to specified host and port
2018-04-26 00:03:46 +02:00
9696953cd7
revise force-restart: now with wait for kill and resource cleanup before start
...
allows to cleanup various hacks, such as checking for pid in vmm_resources
or removing temporarily the allocated resources from the resource map in vmm_engine
semantics is now slightly different, but for sure enhanced.
- each VM has a Lwt.wait () task attached in Vmm_engine.t (tasks : 'c String.Map.t)
- normal create shouldn't be much different, apart from memoizing the sleeper
- after waitpid is done in vmmd, and vmm_engine.shutdown succeeded, Lwt.wakeup is called for the sleeper
- force create now:
- checks static policies
- looks for existing VM (and task), if present: kill and wait for task in vmmd
- continue with presence checking of vm name, dynamic policies, allocate resources (tap, img, fifo)
this means the whole randomness in filenames can be removed, and the
communication between vmm_console and vmm_client is working again (attach/detach
could not work since vmm_console knew only about "albatross.AAA.BBB.RANDOM",
whereas vmm_client insisted on "AAA.BBB"
resource overcommitment (and races in e.g. block device closing + opening) are
gone now, only if the old vm is cleanup up, resources for the new one are
allocated and it is executed
2018-04-05 01:02:45 +02:00
a89b2925fd
Vmm_core.cmd is now a variant (no longer polymorphic variant), some renames in Vmm_wire.Stats and Vmm_wire.Console to disambiguate
2018-04-04 22:16:31 +02:00
fdab43aed6
vmm_engine: finish renaming Destroy_image -> Destroy_vm (broken since bb61388cfc)
2018-04-04 21:46:37 +02:00
7a4661b2e1
style: require lwt 3.0.0, fix warnings, disable 4 (fragile pattern matching) and 48 (implicit elimination of optional argument)
2018-04-03 22:58:31 +02:00
a0c0f39734
vmm_stats: more debug, ignore vmmapi_open failure
...
vmm_stats_lwt: drop all pids on socket disconnect
vmmd: setup statistics slightly later (after the chmod on FreeBSD)
2018-04-01 23:59:12 +02:00
c978dcfb14
less debug output of resources
2018-03-22 23:30:15 +01:00
e48d3b3457
vmm_lwt/vmm_tls: less debug output of read/write stuff
2018-03-22 17:00:08 +01:00
d3941e70c6
style
2018-03-22 17:00:08 +01:00
b9d5fa94f9
call waitpid() earlier, separate freebsd kludge (chmod g+rw), fixes #9
2018-03-22 17:00:08 +01:00
db8ae1ee37
compression, fixes #6
2018-03-22 17:00:08 +01:00
Hannes Mehnert
Stefan Grundmann