Managed instances operations

SSH access

mg --customer $CUSTOMER ssh

Accessing the Docker containers

SSH into the relevant instance and then:

docker ps

You can then use regular Docker commands (e.g. docker exec -it $CONTAINER sh) to interact with the containers.

Accessing the Cloud SQL

mg --customer $CUSTOMER db proxy

or

mg --customer $CUSTOMER db cli

If you find that the command hangs on the following error:

Waiting for cloud_sql_proxy to be ready...

It’s likely that you need to install cloud_sql_proxy.

Restarting for configuration updates

mg --customer $CUSTOMER reload-config

Port-forwarding

mg forward <remote_port> <local_port>

Expose frontend at 8080

mg forward 80 4444

This will port-forward localhost:4444 to port 80 on the VM instance. Some common ports:

• 80: Frontend (also see accessing through the GCP load balancer)
• 3370: Grafana
• 16886: Jaeger

Note that other ports are prevented by the allow-iap-tcp-ingress firewall rule.

Backup

Everything

mg backup

Just the Cloud SQL

mg backup --types sql

Just the VM

mg backup --types vm

Access through the GCP load balancer as a user would

Users access Sourcegraph through GCP Load Balancer/HTTPS -> the Caddy load balancer/HTTP -> the actual sourcegraph-frontend/HTTP. If you suspect that an issue is being introduced by the GCP load balancer itself, e.g. perhaps a request is timing out there due to some misconfiguration, then you will need to access through the GCP load balancer itself. If the managed instance is protected by the load balancer firewall / not publicly accessible on the internet, then it is not possible for you to access $CUSTOMER.sourcegraph.com as a normal user would.

You can workaround this by proxying your internet traffic through the instance itself - which is allowed to reach and go through the public internet -> the GCP load balancer -> back to the instance itself. To do this, create a SOCKS5 proxy tunnel over SSH (replace default-red-instance, if needed):

bash -c '(gcloud beta compute ssh --zone "us-central1-f" --tunnel-through-iap --project "sourcegraph-managed-$CUSTOMER" default-$DEPLOYMENT-instance -- -N -p 22 -D localhost:5000) & sleep 600; kill $!'

Then test you can access it using curl:

$ curl --proxy socks5://localhost:5000 https://$CUSTOMER.sourcegraph.com
<a href="/sign-in?returnTo=%2F">Found</a>.

You can now reproduce the request using curl, or configure your OS or browser to use socks5://localhost:5000:

• Windows: Follow the “using the SOCKS proxy” section in this article [mirror] to enable it on Internet Explorer, Edge and Firefox.
• macOS: System Preferences → Network → Advanced → Proxies → check “SOCKS proxy” and enter the host and the port.
• Linux: Most browsers have proxy settings in their Settings/Preferences.
• Command-line apps: Many CLIs accept http_proxy or https_proxy environment variables or arguments you can set the proxy. Consult the help or the manpage of the program.

IMPORTANT: Once you are finished, terminate the original gcloud beta compute ssh command so your machine’s traffic is no longer going over the instance. The command above will automatically terminate after 10 minutes, to prevent this.

Finding the external IPs

$ gcloud compute addresses list --project=sourcegraph-managed-$CUSTOMER
NAME                     ADDRESS/RANGE   TYPE      PURPOSE  NETWORK  REGION       SUBNET  STATUS
default-global-address   $GLOBAL_IP      EXTERNAL                                         IN_USE
default-nat-manual-ip-0  $NAT_IP_ONE     EXTERNAL                    us-central1          IN_USE
default-nat-manual-ip-1  $NAT_IP_TWO     EXTERNAL                    us-central1          IN_USE

• $GLOBAL_IP is the IP address that $CUSTOMER.sourcegraph.com should point to, it is the address of the GCP Load Balancer.
• $NAT_IP_ONE and $NAT_IP_TWO are the external IPs from which egress traffic from the deployment will originate from. These are the addresses from which Sourcegraph will access the customer’s code host, and as such the customer will need to allow them access to e.g. their internal code host.

Resizing Disks

Disk storage can be safely increased on managed instances at any time. Do not try to decrease the disk size - Terraform will destroy and recreate the disk causing data loss.

To increase the disk size:

1. Set up your variables as usual.

2. Increase the value of data_disk_size in terraform.tfvars and run terraform apply

3. Commit and push your changes:

   git add terraform.tfvars && git commit -m "$CUSTOMER: increase disk size"
   git push origin HEAD
   

4. Follow the GCP instructions to resize the block storage. In most cases, the commands will look like:

   ../util/ssh-exec.sh "sudo resize2fs /dev/sdb"
   

   Then confirm the new size is visible:

   ../util/ssh-exec.sh "df -h /dev/sdb"
   

Running these commands will have no impact on a running deployment and can be safely performed without interruption to the customer.

Capturing network traffic for analysis on the instance

In some cases, you may need to capture network traffic for debugging issues on the instance. We use Wireshark and tcpdump to do this. First, find the service you are interested in, for example, to capture traffic to/from the sourcegraph-frontend service:

  sudo tcpdump -i any -s 65535 'port 3080' -w /tmp/sourcegraph-frontend.pcap

Next you need to scp this from the instance:

   # after eval $(mg workon)
   gcloud compute scp root@default-$DEPLOYMENT-instance:/tmp/sourcegraph-frontend.pcap . # copy from instance

Open the pcap file in Wireshark (installable with brew install --cask wireshark)

Uptime Checks

SOC2/CI-87

We are aligned with the company-wide incident response playbook to handle managed instances downtime.

We utilize GCP Uptime Checks to perform uptime checks against the managed instance frontend url. When such alert is fired, it usually means the service is completely not accessible to customers. In the event of downtime, GCP will notify On-Call DevOps engineers via Opsgenie and the On-Call engineers will proceed with our incident playbook to ensure we reach to a resolution.

Performance Checks

SOC2/CI-25

We utilize the Sourcegraph built-in alerting to monitor application-level metrics. We identify a list of critical metrics that are representation on the overall system performance, and the alert is delivered to Opsgenie. Opsgenie will notify On-Call DevOps engineers](../index.md#on-call) and the On-Call engineers will proceed to investigate and ensure we reach to a resolution.

A list of critial metrics that will be routed to Opsgenie:

• postgres: usage_connections_percentage
• gitserver: disk_space_remaining
• repo-updater: perms_syncer_perms
• repo-updater: perms_syncer_stale_perms
• frontend: 99th_percentile_search_request_duration
• frontend: 99th_percentile_search_codeintel_request_duration

Impact of recreating the instance via Terraform

All configuration about the instance itself is stored in Terraform, so recreating the instance is a non-destructive action. A brand new VM will be provisioned by Terraform, the startup script will initialize it and mount the existing data disk back into the VM, and the Sourcegraph Docker containers will start up.

This typically involves around 8m40s of downtime: 6m destroying the network endpoint group, and 2m creating the new instance / installing software / launching Docker services.

Instance is recreated when startup script changes

Any time a startup script is changed, Terraform will plan to delete the old VM instance and recreate it such that the script runs again.

This is a non-destructive action, aside from the fact that it involves downtime for the deployment.

Debugging startup scripts

View startup script logs

cat /var/log/syslog | grep startup-script

Run startup script and debug:

sudo google_metadata_script_runner --script-type startup --debug

WARNING: Running our startup script twice is a potentially harmful action, as it is usually only ran once.

More details: https://cloud.google.com/compute/docs/startupscript

Viewing container logs

Containers logs are persisted in GCP Logging by utilizing the GCP Logging Driver.

Let’s say you want to check the logs of sourcegraph-frontend-0.

Visit https://console.cloud.google.com/logs and ensure you’re in the right GCP project. Then you may write the following query:

There’s a Show query toggle at the top right corner, turn it on

jsonPayload.container.name : sourcegraph-frontend-0

Learn more about the query language syntax

Fix corrupted repo on gitserver

Context of why this exists:

• #25264
• sourcegraph/customer#887
• sourcegraph/customer#1014 (comment)

A broken repo can be identified by

• Checking https://sourcegraph.example.com/site-admin/repositories?status=failed-fetch
• repo-updater alerts - syncer_synced_repos
• git prune and git fetch operations failing when run inside gitserver on the repo directly

Also possible to fix this by calling a specific API endpoint mutation:

query {
  repository(name: "github.com/sourcegraph/sourcegraph") {
    id
  }
}

mutation {
  recloneRepository(repo: "id-from-query-above") {
    id
  }
}

Once you have identified a repo is constantly failing to be updated/fetched, execute the following steps:

1. Set up env vars

   export PROJECT_PREFIX=sourcegraph-managed
   export CUSTOMER=<customer_or_instance_name>
   export DEPLOYMENT=$(gcloud compute instances list --project "$PROJECT_PREFIX-$CUSTOMER" | grep -v "executors" | awk 'NR>1 { if ($1 ~ "-red-") print "red"; else print "black"; }')
   

2. Determine if git prune or git fetch is failing by exec’ing into the gitserver-0 container

mg ssh
docker exec -it gitserver-0 sh
cd /data/repos/<repo_name>/.git
cat sgm.log
cat gc.log
# look for errors and numbers of failures
# Also run
git prune && git fetch # check for errors

1. Run the following script, from within a clone of sourcegraph/deploy-sourcegraph-managed, to have repo-updater queue an update

   ./util/fix-dirty-repo.sh github.com/org/repo
   

2. Possibly add YAML below per sourcegraph/customer#1128 (comment). This depends on if SRC_ENABLE_SG_MAINTENANCE is thought to be part of the issue.

   gitserver-0:
     environment:
       - SRC_ENABLE_SG_MAINTENANCE=false
       - SRC_ENABLE_GC_AUTO=true
   

Investigate VM platform logs

Navigate to GCP Logging in the right project, use the following query. This is helpful to figure out automated operation against our VM instances.

resource.type="gce_instance"
protoPayload.authenticationInfo.principalEmail="system@google.com"

Enabling GCP Cloud Tracing on Managed Instances

As part of the Centralised Observability efforts we will be enabling GCP Cloud Tracing on all Managed Instances.

To enable GCP Cloud Tracing on an existing instance, follow these instructions for deploying the opentelemetry-collector service:

1. cd into the desired customer directory and into the deployment folder (red or black).
2. Create a new directory called otel-collector and within it a folder called config.yaml with the following contents:

receivers:
  otlp:
    protocols:
      grpc:

exporters:
  jaeger:
    endpoint: '$JAEGER_HOST:14250'
    tls:
      insecure: true
  googlecloud:
    retry_on_failure:
      enabled: false

extensions:
  health_check:
    port: 13133
  zpages:
    endpoint: 'localhost:55679'

service:
  extensions: [health_check, zpages]
  pipelines:
    traces:
      receivers: [otlp]
      exporters: [jaeger, googlecloud]

3. Update the corresponding docker-compose.override.yaml file to now include the otel-collector service and configure all existing services with the correct environment variables. Reference this pull request enabling otel for tpgi for an example of all that needs to be modified.
4. Connect to the manage instance and add the following configuration to the site-config (substituting in the correct GCP Project):

"observability.tracing": {
  "type": "opentelemetry",
  "sampling": "selective",
  "urlTemplate": "https://console.cloud.google.com/traces/list?tid={{ .TraceID }}&project=sourcegraph-managed-$COMPANY"
}

5. Run mg sync artifacts and when complete, restart the instance.
6. Verify on GCP that traces delivered via HTTP POST are now available.

Deploy new images across all instances

Use case: you would like to roll out a new images to all instances

• Open a PR to update the golden file and merge it
• Visit GitHub Actions - reload instances
• Click Run workflow (omit customer slug unless you only want to target a specific customer) and it will run mg sync artifacts then reload deployment on each instance

Update application config across all instances

Use case: you would like to update site-config for all instances

• Open a PR to update mg codes with the right configuration
• Visit GitHub Actions - sync instances config
• Click Run workflow and it will run mg sync on each instance

This action also runs every 24h to ensure all instances config are correct

Make global changes to docker-compose.yaml across all deployment

Use case;

• Configure default cpus of gitserver for all instances
• Use a custom prometheus image for all instances

The docker-compose.yaml of each custom deployment artifacts is a symlink to golden/docker-compose.$version.yaml. The golden file is merged from the upstream file from [sourcegraph/deploy-sourcegraph-docker] and global override files at golden/overrides.

The global override files are separated into different files by purpose. When adding a new override, you should consider whether you should create a new one or append your override in an existing file. Upon updating the global override files, you should run the command below to update the golden files. The update-golden command will pull the target file from upstream, then merge it with every global override files.

NOTES: you should never edit the golden files manually, let mg CLI to do the work

mg update-golden -target $version

Commit your change and make a PR. Once the PR is merged, you can follow this process to roll out changes to all instances.

How to add feature flags?

Feature flags are defined in either config.global.yaml or $CUSTOMER/config.yaml. $CUSTOMER/config.yaml takes higher precedence than config.global.yaml

Clone the repo locally and add your override to the config.global.yaml or $CUSTOMER/config.yaml, then open a PR and ask for review.

Upon merging, follow update application config across all instances.

FAQ: “googleapi: Error 400: The network_endpoint_group resource … is already being used”

If terraform apply is giving you:

Error: Error when reading or editing NetworkEndpointGroup: googleapi: Error 400: The network_endpoint_group resource 'projects/sourcegraph-managed-$COMPANY/zones/us-central1-f/networkEndpointGroups/default-neg' is already being used by 'projects/sourcegraph-managed-$COMPANY/global/backendServices/default-backend-service', resourceInUseByAnotherResource

Or similar—this indicates a bug in Terraform where GCP requires an associated resource to be deleted first and Terraform is trying to delete (or create) that resource in the wrong order.

To workaround the issue, locate the resource in GCP yourself and delete it manually and then terraform apply again.

FAQ: I don’t see an option to log-in as sourcegraph admin

This likely means built-in auth-provider has been disabled. To fix this:

• connect to the DB using mg db
• run SELECT contents FROM critical_and_site_config ORDER BY created_at DESC LIMIT 1; and verify whether the array value of auth.providers contains a provider of type “builtin”
• if there’s no bultin provider, modify the JSON array to add {"type": "builtin","allowSignup":false} and run UPDATE critical_and_site_config SET contents = $JSON where id = (SELECT id FROM critical_and_site_config ORDER BY created_at DESC LIMIT 1) replacing $JSON with the value you modified
• quit mg sql, use mg reload-config to make sure frontend picks-up new changes

FAQ: sourcegraph-admin login credentials are incorrect

This likely means our admin user account was deleted. To verify

• connect to the DB using mg db
• run SELECT id,username,deleted_at FROM users ORDER BY created_at LIMIT 1; and check if the third value is set to a non NULL value (a date)
• if you see deleted_at set, take note of ID (usually equal to 1, returned in first column) and use UPDATE users SET deleted_at = NULL where ID = ? (substitute ? for id) to mark the user as not deleted
• login should now work correctly