def _get_kube_config(self):
if j.sals.vdc.list_all():
self.vdc_name = list(j.sals.vdc.list_all())[0]
else:
raise StopChatFlow(f"No Virtual Data Centers(VDC) were found.",
htmlAlert=True)
self.vdc_info = {}
self.vdc = j.sals.vdc.find(name=self.vdc_name, load_info=True)
self.identity_name = j.core.identity.me.instance_name
self.secret = f"{self.vdc.identity_tid}:{uuid.uuid4().hex}"
for node in self.vdc.kubernetes:
if node.role == KubernetesRole.MASTER:
self.vdc_info["master_ip"] = node.ip_address
self.vdc_info["pool_id"] = node.pool_id
self.vdc_info["public_ip"] = node.public_ip
self.vdc_info[
"farm_name"] = j.core.identity.me.explorer.farms.get(
j.core.identity.me.explorer.nodes.get(
node.node_id).farm_id).name
self.vdc_info["kube_config_path"] = j.sals.fs.expanduser(
"~/.kube/config")
self.vdc_info["network_name"] = self.vdc_name
self.vdc_info["network_view"] = deployer.get_network_view(
self.vdc_name, identity_name=self.identity_name)
break
def _get_pool(self):
self._get_available_farms(only_one=False,
identity_name=self.identity_name)
self._select_farms()
self._select_pool_node()
self.pool_info = deployer.create_3bot_pool(
self.farm_name,
self.expiration,
currency=self.currency,
identity_name=self.identity_name,
**self.query,
)
if self.pool_info.escrow_information.address.strip() == "":
raise StopChatFlow(
f"provisioning the pool, invalid escrow information probably caused by a misconfigured, pool creation request was {self.pool_info}"
)
payment_info = deployer.pay_for_pool(self.pool_info)
result = deployer.wait_pool_reservation(self.pool_info.reservation_id,
bot=self)
if not result:
raise StopChatFlow(
f"provisioning the pool timed out. pool_id: {self.pool_info.reservation_id}"
)
self.md_show_update(
f"Capacity pool {self.pool_info.reservation_id} created and funded with {payment_info['total_amount_dec']} TFT"
)
gevent.sleep(2)
self.wgcfg = deployer.init_new_user_network(
self,
self.identity_name,
self.pool_info.reservation_id,
identity_name=self.identity_name,
network_name="management",
)
self.md_show_update("Management network created.")
self.pool_id = self.pool_info.reservation_id
self.network_view = deployer.get_network_view(
"management", identity_name=self.identity_name)
def redeploy_threebot_solution(
owner,
solution_uuid,
backup_password,
compute_pool_id=None,
gateway_pool_id=None,
solution_info=None,
node_id=None,
bot=None,
retry=False,
prompt_retry_only=True,
):
"""
Args:
owner (str): threebot_name of the logged in user
solution_uuid (str): of the not-running threebot that needs to be started
compute_pool_id (str): to override the pool id used for container deployment. if not specified, it will use the old pool id
gateway_pool_id (str): to override the pool id used for subdomain and proxy deployment. if not specified, it will use the old pool id. (should override the subdomain specified in solution_info)
solution_info (dict): to override the information used in deployment. if any key is not specified, it will use the old value
"""
if prompt_retry_only:
msg_bot = None
else:
msg_bot = bot
retries = 3 if retry else 1
solution_info = solution_info or {}
owner = text.removesuffix(owner, ".3bot")
j.logger.debug(f"Fetching solution info for uuid {solution_uuid} and owner {owner}")
while retries:
try:
if bot:
bot.md_show_update("Starting your 3Bot...")
threebot = get_threebot_config_instance(owner, solution_uuid)
if not threebot.verify_secret(backup_password):
raise j.exceptions.Validation(f"incorrect secret provided")
zos = get_threebot_zos(threebot)
identity = generate_user_identity(threebot, backup_password, zos)
with threebot_identity_context(identity.instance_name):
j.logger.debug(f"Using identity {identity.instance_name}")
with deployment_context():
zos = j.sals.zos.get(identity.instance_name)
solution_workloads = get_threebot_workloads_by_uuid(solution_uuid, identity.instance_name)
new_solution_info = build_solution_info(solution_workloads, threebot)
j.logger.debug(f"old solution info: {new_solution_info}")
new_solution_info.update(solution_info)
j.logger.debug(f"updated solution info: {new_solution_info}")
gateway_pool_id = gateway_pool_id or new_solution_info["gateway_pool"]
compute_pool_id = compute_pool_id or new_solution_info["compute_pool"]
# deploy using the new information with a new uuid. a new uuid to not conflict with the old one when listing
solution_name = new_solution_info["name"]
backup_model = BACKUP_MODEL_FACTORY.get(f"{solution_name}_{owner}")
new_solution_uuid = uuid.uuid4().hex
metadata = {
"form_info": {"Solution name": solution_name, "chatflow": "threebot"},
"owner": f"{owner}.3bot",
}
# select node and update network
j.logger.debug(f"fetching network {new_solution_info['network']}")
network_view = deployer.get_network_view(
new_solution_info["network"], identity_name=identity.instance_name
)
j.logger.debug(f"searching for available node within pool {compute_pool_id}")
if node_id:
selected_node = zos._explorer.nodes.get(node_id)
else:
selected_node = deployer.schedule_container(
pool_id=compute_pool_id,
cru=new_solution_info["cpu"] + 1,
mru=(new_solution_info["memory"] / 1024) + 1,
sru=(new_solution_info["disk_size"] / 1024) + 0.25,
ip_version="IPv6",
)
j.logger.debug(f"found node with enough capacity {selected_node.node_id}")
j.logger.debug(f"adding node {selected_node.node_id} to network {network_view.name}")
result = deployer.add_network_node(
network_view.name,
selected_node,
compute_pool_id,
network_view,
bot=msg_bot,
identity_name=identity.instance_name,
)
if result:
for wid in result["ids"]:
success = deployer.wait_workload(
wid,
msg_bot,
breaking_node_id=selected_node.node_id,
identity_name=identity.instance_name,
)
if not success:
raise DeploymentFailed(
f"Failed to add node {selected_node.node_id} to network {wid}", wid=wid
)
j.logger.debug(f"node {selected_node.node_id} added to network {network_view.name} successfuly")
j.logger.debug("searching for a free ip address")
network_view = network_view.copy()
ip_address = network_view.get_free_ip(selected_node)
j.logger.debug(f"found a free ip address {ip_address}")
workload_ids = []
j.logger.debug(f"fetching gateway {new_solution_info['gateway']}")
gateway = zos._explorer.gateway.get(new_solution_info["gateway"])
addresses = []
j.logger.debug(f"resolving gateway {gateway.node_id} name servers")
for ns in gateway.dns_nameserver:
try:
addresses.append(j.sals.nettools.get_host_by_name(ns))
except:
j.logger.error(f"failed to resolve name server {ns} of gateway {gateway.node_id}")
if not addresses:
raise StopChatFlow(
f"the gateway specfied {gateway.node_id} doesn't have any valid name servers"
)
domain = new_solution_info["domain"]
j.logger.debug(f"deploying domain {domain} pointing to addresses {addresses}")
workload_ids.append(
deployer.create_subdomain(
pool_id=gateway_pool_id,
gateway_id=gateway.node_id,
subdomain=domain,
addresses=addresses,
solution_uuid=new_solution_uuid,
identity_name=identity.instance_name,
**metadata,
)
)
j.logger.debug(f"waiting for domain workload {workload_ids[-1]} to deploy")
success = deployer.wait_workload(
workload_ids[-1], bot=msg_bot, identity_name=identity.instance_name
)
if not success:
raise DeploymentFailed(
f"Failed to create subdomain {domain} on gateway {gateway.node_id} {workload_ids[-1]}. The resources you paid for will be re-used in your upcoming deployments.",
wid=workload_ids[-1],
identity_name=identity.instance_name,
)
test_cert = j.config.get("TEST_CERT")
j.logger.debug("creating backup token")
backup_token = str(j.data.idgenerator.idgenerator.uuid.uuid4())
backup_model.token = backup_token
backup_model.tname = metadata["owner"]
backup_model.save()
j.logger.debug(f"backup token {backup_token} created for tname {backup_model.tname}")
environment_vars = {
"SDK_VERSION": new_solution_info["branch"],
"INSTANCE_NAME": new_solution_info["name"],
"THREEBOT_NAME": owner,
"DOMAIN": domain,
"SSHKEY": new_solution_info["public_key"],
"TEST_CERT": "true" if test_cert else "false",
"MARKETPLACE_URL": f"https://{j.sals.nginx.main.websites.threebot_deployer_threebot_deployer_root_proxy_443.domain}/",
"DEFAULT_IDENTITY": "test" if "test" in j.core.identity.me.explorer_url else "main",
}
j.logger.debug(f"deploying threebot container with environment {environment_vars}")
log_config = j.core.config.get("LOGGING_SINK", {})
if log_config:
log_config["channel_name"] = f'{owner}-{new_solution_info["name"]}'.lower()
workload_ids.append(
deployer.deploy_container(
pool_id=compute_pool_id,
node_id=selected_node.node_id,
network_name=network_view.name,
ip_address=ip_address,
flist=new_solution_info["flist"],
env=environment_vars,
cpu=new_solution_info["cpu"],
memory=new_solution_info["memory"],
disk_size=new_solution_info["disk_size"],
secret_env={"BACKUP_PASSWORD": backup_password, "BACKUP_TOKEN": backup_token},
interactive=False,
log_config=log_config,
solution_uuid=new_solution_uuid,
identity_name=identity.instance_name,
**metadata,
)
)
j.logger.debug(f"wating for threebot container workload {workload_ids[-1]} to be deployed")
success = deployer.wait_workload(
workload_ids[-1], bot=msg_bot, identity_name=identity.instance_name
)
if not success:
raise DeploymentFailed(
f"Failed to create container on node {selected_node.node_id} {workload_ids[-1]}. The resources you paid for will be re-used in your upcoming deployments.",
solution_uuid=new_solution_uuid,
wid=workload_ids[-1],
identity_name=identity.instance_name,
)
j.logger.debug(f"threebot container workload {workload_ids[-1]} deployed successfuly")
trc_log_config = j.core.config.get("LOGGING_SINK", {})
if trc_log_config:
trc_log_config["channel_name"] = f'{owner}-{new_solution_info["name"]}-trc'.lower()
identity_tid = identity.tid
secret = f"{identity_tid}:{uuid.uuid4().hex}"
j.logger.debug(f"deploying trc container")
workload_ids.extend(
deployer.expose_address(
pool_id=compute_pool_id,
gateway_id=gateway.node_id,
network_name=network_view.name,
local_ip=ip_address,
port=80,
tls_port=443,
trc_secret=secret,
node_id=selected_node.node_id,
reserve_proxy=True,
domain_name=domain,
proxy_pool_id=gateway_pool_id,
solution_uuid=new_solution_uuid,
log_config=trc_log_config,
identity_name=identity.instance_name,
**metadata,
)
)
j.logger.debug(f"wating for trc container workload {workload_ids[-1]} to be deployed")
success = deployer.wait_workload(
workload_ids[-1], bot=msg_bot, identity_name=identity.instance_name
)
if not success:
raise DeploymentFailed(
f"Failed to create TRC container on node {selected_node.node_id} {workload_ids[-1]}. The resources you paid for will be re-used in your upcoming deployments.",
solution_uuid=new_solution_uuid,
wid=workload_ids[-1],
identity_name=identity.instance_name,
)
j.logger.debug(f"trc container workload {workload_ids[-1]} deployed successfuly")
j.logger.debug(f"fetching farm information of pool {compute_pool_id}")
farm_id = deployer.get_pool_farm_id(compute_pool_id)
farm = zos._explorer.farms.get(farm_id)
j.logger.debug(f"saving new threebot local config with uuid {new_solution_uuid}")
instance_name = f"threebot_{new_solution_uuid}"
user_threebot = USER_THREEBOT_FACTORY.get(instance_name)
user_threebot.solution_uuid = new_solution_uuid
user_threebot.identity_tid = identity.tid
user_threebot.name = solution_name
user_threebot.owner_tname = threebot.owner_tname
user_threebot.farm_name = farm.name
user_threebot.state = ThreebotState.RUNNING
user_threebot.continent = farm.location.continent
user_threebot.explorer_url = identity.explorer_url
user_threebot.subdomain_wid = workload_ids[-4]
user_threebot.threebot_container_wid = workload_ids[-3]
user_threebot.trc_container_wid = workload_ids[-2]
user_threebot.reverse_proxy_wid = workload_ids[-1]
user_threebot.save()
j.logger.debug(f"threebot local config of uuid {new_solution_uuid} saved")
j.logger.debug(f"deleting old threebot local config with uuid {solution_uuid}")
USER_THREEBOT_FACTORY.delete(f"threebot_{solution_uuid}")
j.logger.debug("deployment successful")
return user_threebot
except DeploymentFailed as e:
retries -= 1
if retries > 0:
j.logger.error(f"3Bot {solution_uuid} redeployment failed. retrying {retries}")
if bot and e.wid:
bot.md_show_update(f"Deployment Failed for wid {e.wid}. retrying {retries} ....")