Bugfix release

Fix for problem introduced in 2.4.5.1 See: #1910

Release v2.4.5.1

Features:

FR: close socks proxy on ROS upon ROS creation

Issues:

Statscollector doesn't register new node after start

Release: v2.4.5

Features:

• Add multiple nic interfaces on a vm
• Mass VM creation test case
• Update version page
• Automate release process
• Auto push latest versions of OS images to G8's in the field
• Ability to create different VLANs

Issues:

• When enabling a node that was previous in maintenance not all healthchecks are rescheduled
• Selfheal routeros when halted routeros detected
• Can't delete CS from "enduser portal"
• AttributeError: 'cloudbroker_externalnetwork' object has no attribute 'pingsips'
• Healthcheck for ECO's
• Error displaying job page
• Snapshot rollback doesn't work for data disk
• Error when attaching a newly created external network to a running vm.
• Cannot type the reason of deleting vm while opening console form CB portal
• ResponseError: Error running script (call to f_2e05a33aa8d06be63375e7e11c1131334863c42e): @user_script:99: user_script:99: Cannot serialise number: must not be NaN or Inf in greenitglobe_ovs_vdisks
• Error in jumpscr:greenitglobe_disk_garbage_collector
• AttributeError: 'NoneType' object has no attribute 'find' jumpscale_disk_virtual_stats
• Cannot place node halted ( crashed / frozen ) in maintenance
• Remove js9 dependency
• Improve error handling whilst upgrading

Bugfix + Minor features

Notes:

This release contains small feature and bugfixes.

Update process:

Due to change #1619 some manual updates for the system-config are required prior to running this update.

You will need to update the system-config,yaml to have the ssl certificates embedded see example. Take care to make sure you give the certificates the same name as before the update.

An easy way to accomplish this is by running a script on one of the controllers

wget https://raw.githubusercontent.com/0-complexity/openvcloud_installer/d57d2f1cc7114171226e97211450f00a5c06208f/scripts/migrations/migrate244
apt-get install python3-yaml
python3 migrate244

And then run the normal update.

Alternatively you can change it manually and write the system-config,
Then run the regular update.

Make sure to update the system-config in the git repository aswell.

Features:

• FR: Add recycle bin for OVC resources
• FR: Make it possible to access vm console from admin portal
• Redirect console to logfile
• Build full docker images
• Nightly builds & testruns
• ROSes should be distributed on all enabled cpu nodes
• FR: create a new option to delete the VM with its disks immediately without waiting retention period
• Add delete option for the CD-Rom images
• FR - when deleting VMs, make sure no orphanage is
• add information regarding any running jobs for all elements (VDC/VM/Stack)

Bug fixes:

• Refresh health in portal does not work
• Deployment test cannot succeed cause VMs are limited to 2000 iops
• disk orphan process is crashing and vdisks are not cleared up.
• Invalid response content when specifying a wrong authkey in an API request
• Failing to import dsl.ova image
• export on specific machine fails without email notification or error being displayed on the portal
• Remove/move disk doesn't work from user portal on be-g8-3
• Remove the At your service tab from in the admin portal
• Define on image level wether a vm supports hot-add cpu/mem
• SSL certificate should be part of the system-config.yaml
• 0-access session leakage
• Cannot remove IP from Free ip addresses list (external network)
• Revise firewall rules on controller nodes
• Recycle bin autotests
• Internal server error when opening certain ecos
• Restoring VM does not check resource limiting
• Need to take locks on account and cloudspace level when deploying new assets
• openvcloud_installler: don't use "sender" field of system-config
• Removing Orphan OVS bridges
• bug - adding pub key to IYO profile is compulsory for "normal users"
• store node shown as in maintenance mode on js shell while it is shown as enabled in portal
• Generated reports from Resource Consumption Tracking does not work properly
• No data in ACL User portal
• API call to "cloudapi/users/authenticate" at DS1/MR4 does not work
• LimitIO for Disk errors vaguely for negative value
• Deployment_test : Failed to create VM
• 2.4.0 kubernetes setup doesn't include ipmi entries for the controllers
• ROS crashing under high load
• OVC 2.3.0 GRAFANA is not working properly
• Failed executing job heartbeat check / Jobs Section shows us nothing
• Generate Billing Records for Individual Accounts
• FR: Admin on cloudspace should be able to list/get the cloudspace account
• Need to make this healtcheck [ volumedriver_check.py ] more smart incase a volumedriver actually is down
• Generate notification when Storage node is placed in maintenance
• Make sure that redis ins't exposed in production mode
• OVC request without Accept request header results in text/plain response type
• When an orphan disk that has clones gets deleted it crashes the entire orphan

Bug fix release

• disk type filter in /cloudapi/disks/list doesn't work
• Extending the automated tests
• Machine ID redirection is not working propely from VDisk info page
• vm creations fails when account has resourcelimits
• VM migration failed due to host key checking with node itself
• Enable storage node failed in ovc 2.4.2
• FR: Change the alarm generated on alerta from performance tests to values smaller than 10Mbps
• timeout increasing for healthcheck script routeros_check_schedule.py up to 3 minutes
• vdisks are not deleted in OVS after the retention period on OVC
• Add redirection from https://portal.domain/g8vdc -> https://portal.domain/g8vdc/ to nginx config in next installations
• Network Address Validation Prior to Database insertion
• enable specific account id option for the CD-Rom images
• Access error when trying to delete disk

Feature Release

• portal returns error while opening image page
• create vm on specific node returns error
• Delete snapshot from api doesn't work successfully
• can't migrate vm
• POST /cloudbroker/machine/attachExternalNetwork returns internal server error
• Adds allow creating portforwading with different protocol for same port
• 2.3.0 vm export doesn't work on second attempt
• images does not removed auto from /opt after creating an images on OVC
• Better error message when out of storage
• Memory Index shows NaN
• disabled image appear in end user portal
• POST /cloudapi/disks/delete returns internal server error
• /cloudapi/machines/clone returns internal server error
• Routeros healtcheck doesn't detect crashed ros
• Can't re-start machines
• OVC 2.3.0 VDC stuck in deleting status
• OVC 2.3.0 Accessing advance Shield configuration error on password
• OVC 2.3.0 Port forward stuck on OK button
• OVC 2.3.0 Failed to delete port forward error, while delete was successful
• OVC 2.3.0 Changing port forward returns bogus message about VM
• OVC 2.3.0 Migration / upgrade script is not working
• OVC 2.3.0 duplicated packages during migration
• OVC 2.3.0 cannot delete cloud space even when it doesn't contain any VMs
• FR: Add Alerta Configuration to the installation script of ovc
• Grid last checked is not being displayed on most nodes
• MASQUERADE role missing in installing ovc
• Use moveVirtualFirewallToFirewallNode API with non-exist node return 500 instead of 404
• Got error on Deactive a stor node from portal
• 0-access terminates ssh connections in arbitrary time
• Don't rely on btrfs for routeros images
• Putting node on maintenance mode : error comes up when move a ROS to a destination cpu, the destination CPU already has an exist snapshot
• Deploy and move VFW for destroyed cloudspace return 200
• Destroy deleted cloudspace return 200 instead of 404
• Can't create vm on cloudspace by user with user group only but has write and admin access to this cloudspace
• Can Create more than one cloudspace on account with ip limitation equal only 1
• Disable CS doesn't disable the router OS
• Create the possibility to have any size of VM's (CPU/MEM) in a variable way not fixed to a size ID.
• id_rsa, id_rsa.pub and known_hosts files in the nodes aren't security wise approach
• i still can access 0-access page using the url with account which is not in itsyouonlie 0-access group
• i can create a vm with image that is not exposed to my account
• /cloudapi/cloudspaces/delete returns 409 even if the cloudspace has no vms
• loghandling issue
• agentcontroller raise an error
• Add ExtraIP and remove IP API doesn't do anything
• need to set max width/height to the portal logo
• Delete non-existing group returns internal server error
• need a sufficient portal messages while putting a node in maintenance to report the error
• VMs/ROS do not move automatically when cpu node rebooted without putting on maintenance mode
• we can update destroyed account name
• Creating a machine with a name of any previously deleted machines fails on be-g8-4
• cloudapi/account/create returns internal server error
• Cannot create port forward for different protocol on the same port number.
• Make it possible to migrate to kubernetes
• Could not execute jscript:178 greenitglobe_ovs_disk_safety
• Online upgrades firmware-bios-code (rolling upgrade)
• Cloudspace migration between environments
• Test case:Add a new test case to cover swap consuming issue
• jumpscript out of frame
• rollbackSnapshot en deleteSnapshot always return true...

Feature Release

Release notes:

New features:

• Master install is based on kubernetes and is setup HA by default.
• Ability to import CD-ROM images and VM tempaltes from URL
• Automates upgrade procedure
• Ability to execute jumpscripts from portal
• Nodes in maintenance don't generate alerts
• VM delete deletes actually data after 7days allowing the possiblity to recover a VM within those 7days
• Cloudspace subnet is configurable during creation of cloudspace
• Allow ability to boot VMs in UEFI mode. (based on the image type)
• Create template from an existing VM (restored functionality)
• Make it possible to boot VM from CD-ROM image (recovery mode)
• Add 0-access (secure login to physical infrastructure)

OpenvCloud issues resolves:

• error when update the environment
• "Virtual Machines which use this image" table always empty
• put node in maintenance fails
• Could not execute macro drawgraph inside cpu page
• Error Condition not found
• RouterOS - configuration rollback
• POST /cloudbroker/image/createCDROMImage raise internal server error
• Zero-os vm
• ROS'es do not migrate as per User Selection
• CloudAPI accepts expired JWTs
• Automated OVS installation procedure
• Create rescue mode for all VMs
• Issue with maintenance mode, not all vms are moved when vm is imported
• Macro error on user page ,if it'syouonline user has a space in his name
• Create upgrade pod
• Create upgrade nginx
• Create upgrade script for pods
• Create upgrade script for pnodes
• When cpu node hard reboots ros on that node are not deletable
• Fix bond name in networkbond_check script
• Internal server error, If I pass something wrong or expired jwt on zeroaccess/provision API
• Can't connect to cntrl2&3 with zero-access
• JWT should be auto-refresh in environments
• VFW moves to a wrong stack other than the assigned one.
• Vfw will always get started when enabling a node.
• Need to stop Alba warning about disks for a node with status 'MAINTENANCE'
• 'CloudBroker' object has no attribute 'getBestProvider'
• ovs_credentials raise an error
• Couldn't start a vm because of that eco and it's irreducable.
• Arakoon Collapse hitting timeout
• VM Creation Failure - Make volume with API first and then write data on it
• user api access to OVC
• typo in 0-access session a
• Secure environment setup
• Fix redischeck when agent is pod
• ays_process_check gives error on portal and agentcontoller
• GET /cloudapi/accounts/getConsumption returns empty zip file
• Allow the ability to change the DHCP IP Address range assigned to VDCs
• need api to get private network info
• Image availability set on account level
• Need to change info in 0-access node player page to be name of node
• Wrong doc string for zeroaccess/getSessionInitTime API
• B17: VMs create double snapshots
• VM cannot obtain IP address from DHCP
• Some weird feature in portal make portal crash
• Disk orphan schedule does not detect problem when disk orphan fails
• When an orphan disk that has clones gets deleted it crashes the entire orphan
• Panel Grid shows VM being on different cpu nodes
• OSIS high cpu time and memory allocation
• Ubuntu pod for OVS installation
• We need to resolve recurring issues ( multiple ECOs of the same condition )
• add asciinma player to portal
• Need health check that monitors for flooding of ECO's
• Could not execute macro processchecks on page status overview
• 0-access pod
• Do not replace custom customer logo with OVC upgrade
• Imported disks do not appear to have names
• Restrict access to 0-access group
• add 0-access page to ovc for a single node
• add 0-access page in ovc portal for all nodes
• Improve metadataiso storagerouter selection
• Create kubernetes cluster agent
• Resize with ubuntu 14.04 doesn't work correctly
• Add VM name to the email message that is sent when exporting/importing completes.
• Mem value is wrong
• Try to restart agent if it dies
• can't migrate vm
• Provide the ability from the panel to pause all vm in the whole environment
• Need api to get external network information
• Automate boot cycle of nodes.
• Automatically create pull request to the relevant git repository with changes made on the code through the portal
• Schedule self healing script from portal
• Security/privacy issue: you can search usernames in the OVC End User Portal
• Add timeout before complete VM deletion after destroy button is pressed
• Add StatefulSets to the kubernetes client
• Automatic upgrade procedure
• Kubernetes cluster Service
• (research/make) PXE boot containers redundant with special network settings
• Create controller aio docker image using prefab for K8s setup
• Create base (jumpscale7/ovc installation) docker image using prefab
• Need redundant storage system for billing capnp files (research/implementation)
• Find a way to share master node config files (hrd) in K8s
• ResetPasswordLink doesn't work correctly
• Automate the change of URL-name for the cloud broker portal
• [Need selfhealing script that scrubs disks if their stored size grows bigge...

Bug fix release

Openvcloud Repo

• Putting node, which has halted VFW, in maintenance with option "move vms" fails
• VFW can be started on a node in maintennce
• Putting node, which has halted VFW, in maintenance with option "stop vms" fails
• Can't stop VFW after putting node in maintenance or after moving vfw.
• Can't start a newly created VFW
• VFW can be moved to a node in maintenance.
• Can't put node in maintenance (with option move vms) which has vm that is connected to an external network.
• A virtual machine can be moved to a cpu-node which is in maintenance using api call.
• OVC should select a second storage router node to host new VM if the primary storage router node failed
• Can't migrate a paused machine.
• VM with 2 different ids.
• Portal page load race condition
• Sync alert to alerta
• VMs do not automatically move when Node in Error or Maintenance
• Need to record audit record when a vm gets shutdown from inside the OS
• Keep audit records for more than 5 days
• cloudapi_machines_listSnapshots fails when snapshot name is number instead of string
• Status Code filter doesn't work in Audits table of Account
• internal server error on logout in VDC
• Add POST /cloudapi/users/create
• Script bug on volumedriver_check.py on be-gen-1
• /cloudapi/account/delete produce 500 error as a status code

Selfhealing Repo:

• when smartctl hangs, after a while 1000's of smartctl -A /dev/$disk ruin the system

Bug fix release

Fixed issues

• Libvirtlistener is not installed during production install
• cloned vm come with two metadata disks
• Harden XML Parsing inside OVC
• osisServerStart.py and controller.py eat server ram
• get machine info returns internal server error

Bug fix release

Fixes issues

• Hanging of portal page when moving an ROS from portal
• Show MTU value in healthcheck
• changing the IO params of a disk does not work
• don't count on timestamp of snapshots
• Suggestion to remove interrupt from the healtchecks
• Configure JS agent in systemd
• Log handling fails
• shouldn't be able to detach the boot disk from a running vm
• Password copy/paste does not work within the Terminal Window of Ubuntu 16.04 Desktop Edition
• "Swap used value" is causing red alarms all over the store nodes with no actual reason.