Marcos F

My feedback

  1. 75 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Facebook Google
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    3 comments  ·  Compute Engine  ·  Flag idea as inappropriate…  ·  Admin →
    Marcos F supported this idea  · 
  2. 232 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Facebook Google
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    13 comments  ·  Compute Engine  ·  Flag idea as inappropriate…  ·  Admin →
    Marcos F commented  · 

    THIS. I work with a project where I had to setup a VM just for scheduling purposes, This would add great value for us.

    Marcos F supported this idea  · 
  3. 200 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Facebook Google
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    under review  ·  11 comments  ·  Compute Engine  ·  Flag idea as inappropriate…  ·  Admin →
  4. 1,183 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Facebook Google
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    65 comments  ·  Compute Engine  ·  Flag idea as inappropriate…  ·  Admin →

    Hi folks, we know there is a ton of interest in this ability, and our engineering team is working on enabling it. We don’t have specific schedule updates right now, but I wanted to post an update acknowledging that we definitely agree this is an important ability and we will launch the functionality just as soon as we can!

  5. 49 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Facebook Google
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    10 comments  ·  Compute Engine  ·  Flag idea as inappropriate…  ·  Admin →

    Hello all. We have now deployed column configuration options in the UI for the GCE area of our Console. We recognize that there have been several requests for additional columns, and we’re happy to implement those. If you’d like to suggest specific columns, please comment on this feature request and we’ll get to work on them! :)

  6. 9 votes
    Vote
    Sign in
    (thinking…)
    Sign in with: Facebook Google
    Signed in as (Sign out)
    You have left! (?) (thinking…)
    under review  ·  2 comments  ·  Compute Engine  ·  Flag idea as inappropriate…  ·  Admin →
    Marcos F commented  · 

    Hey Scott, thanks for jumping in.

    In GCE, we got 27 Windows VM's, with 26 in production (one scheduler, 25 for classes), serving an ERP instance for our students. Each unit will have up to 5 different classes, and we have assigned one external IP for each unit. Each class will have a different VM to keep all the data separated from others and to make possible to return to day zero with the initial snapshot.
    The problem is, sometimes a VM will start up with the external IP resource with the name "external-nat" , and others it will go up with "External NAT", as below (look for "kind": "compute#accessConfig", "type": "ONE_TO_ONE_NAT"):

    {
    "kind": "compute#instance",
    "id": "1349325344426716703",
    "creationTimestamp": "2015-08-20T07:57:17.232-07:00",
    "zone": "https://www.googleapis.com/compute/v1/projects/formal-era-102118/zones/us-central1-f",
    "status": "RUNNING",
    "name": "agendador-protheus11",
    "description": "",
    "tags": {
    "items": [
    "http-server",
    "https-server"
    ],
    "fingerprint": "6smc4R4d39I="
    },
    "machineType": "https://www.googleapis.com/compute/v1/projects/formal-era-102118/zones/us-central1-f/machineTypes/g1-small",
    "canIpForward": false,
    "networkInterfaces": [
    {
    "network": "https://www.googleapis.com/compute/v1/projects/formal-era-102118/global/networks/default",
    "networkIP": "10.240.150.23",
    "name": "nic0",
    "accessConfigs": [
    {
    "kind": "compute#accessConfig",
    "type": "ONE_TO_ONE_NAT",
    "name": "External NAT",
    "natIP": "104.197.67.217"
    }
    ]
    }
    ],
    <edited for space reasons...>}

    And this one:

    {
    "kind": "compute#instance",
    "id": "7446217102326975496",
    "creationTimestamp": "2015-08-11T12:57:32.934-07:00",
    "zone": "https://www.googleapis.com/compute/v1/projects/formal-era-102118/zones/us-central1-b",
    "status": "RUNNING",
    "name": "sao-diadema-turma02",
    "description": "",
    "tags": {
    "items": [
    "sao-diadema",
    "no-ip"
    ],
    "fingerprint": "ablSR86Mi3w="
    },
    "machineType": "https://www.googleapis.com/compute/v1/projects/formal-era-102118/zones/us-central1-b/machineTypes/n1-standard-2",
    "canIpForward": false,
    "networkInterfaces": [
    {
    "network": "https://www.googleapis.com/compute/v1/projects/formal-era-102118/global/networks/default",
    "networkIP": "10.240.64.148",
    "name": "nic0",
    "accessConfigs": [
    {
    "kind": "compute#accessConfig",
    "type": "ONE_TO_ONE_NAT",
    "name": "external-nat",
    "natIP": "162.222.177.241"
    }
    ]
    }
    ],
    <edited for space reasons...>}

    We use Google Cloud SDK scripts to control the behaviour of our VMs clicling in a single external IP, allocating an disallocating it to a new VM when its time for new class to go up like this:

    gcloud compute instances delete-access-config %vm_name% --access-config-name "external-nat" --zone %vm_zone%

    gcloud compute instances add-access-config %vm_name% --address %vm_ip% --zone %vm_zone%

    As a workaround, we had to include another line as follows:

    gcloud compute instances delete-access-config %vm_name% --access-config-name "External NAT" --zone %vm_zone%

    So, there are two different strings for the same kind of resource in VMs in GCE.
    We had a lot of failures during development time and testing wuth this problem and the workaround kind of fixed it. The problem is: GC-SDK commands take some time to execute (from 5 to 45 seconds sometimes) and we had to add one more command to make everithing happen in a more reliable way. Multipiply that by the number of units we have today, and it will start do delay classes in some minutes.

    I've seen at least another fellow having this same problem with this behaviour, wich conducted us to the workaround.

    I hope this clarifies the issue. Thanks again for looking at it.

    Marcos F shared this idea  · 

Feedback and Knowledge Base