github/container.training
1
0
Fork 0
mirror of https://github.com/jpetazzo/container.training.git synced 2026-04-19 00:26:41 +00:00
Code Issues Packages Projects Releases Wiki Activity

Rolling upgrades

Browse Source
This commit is contained in:
Jerome Petazzoni
2016-08-22 14:21:00 -07:00
parent 2c8664e58d
commit 40e0c96c91
1 changed files with 29 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

37
www/htdocs/index.html
View File

@@ -1933,7 +1933,10 @@ It is a virtual IP address (VIP) for the `rng` service.
]
It doesn't respond to ping at this point. (This might change in the future.)
It *should* ping. (But this might change in the future.)
Current behavior for VIPs is to ping when there is a backend available on the same machine.
(Again: this might change in the future.)
---
@@ -2150,7 +2153,7 @@ WHY?!?
- Check that they're running:
```bash
docker service tasks worker
docker service ps worker
```
]
@@ -2202,16 +2205,17 @@ Note how the build and push were fast (because caching).
- Look at our service status:
```bash
watch -n1 "docker service tasks worker -a | grep -v Shutdown.*Shutdown"
watch -n1 "docker service ps worker | grep -v Shutdown.*Shutdown"
```
]
- `-a` gives us all tasks, including the one whose current or desired state is `Shutdown`
- `docker service ps worker` gives us all tasks
<br/>(including the one whose current or desired state is `Shutdown`)
- Then we filter out the tasks whose current **and** desired state is `Shutdown`
- Future versions will have fancy filters to make that less tinkerish
- Future versions might have fancy filters to make that less tinkerish
---
@@ -2228,9 +2232,7 @@ Note how the build and push were fast (because caching).
]
SwarmKit updates all instances at the same time.
If only we could do a rolling upgrade!
By default, SwarmKit does a rolling upgrade, one instance at a time.
---
@@ -2256,6 +2258,25 @@ If only we could do a rolling upgrade!
---
## Timeline of an upgrade
- SwarmKit will upgrade N instances at a time
<br/>(following the `update-parallelism` parameter)
- New tasks are created, and their desired state is set to `Ready`
<br/>.small[(this pulls the image if necessary, ensures resource availability, creates the container ... without starting it)]
- If the new tasks fail to get to `Ready` state, go back to the previous step
<br/>.small[(SwarmKit will try again and again, until the situation is addressed or desired state is updated)]
- When the new tasks are `Ready`, it sets the old tasks desired state to `Shutdown`
- When the old tasks are `Shutdown`, it starts the new tasks
- Then it waits for the `update-delay`, and continues with the next batch of instances
---
## Getting cluster-wide task information
- The Docker API doesn't expose this directly (yet)