Dropping from a cluster drops from every node of the cluster.
Including nodes that the cluster does not think have the content.
This is different from GET and CHECKPRESENT, which do trust the
cluster's location log. The difference is that removing from a cluster
should make 100% the content is gone from every node. So doing extra
work is ok. Compare with CHECKPRESENT where checking every node could
make it very expensive, and the worst that can happen in a false
negative is extra work being done.
Extended the P2P protocol with FAILURE-PLUS to handle the case where a
drop from one node succeeds, but a drop from another node fails. In that
case the entire cluster drop has failed.
Note that SUCCESS-PLUS is returned when dropping from a proxied remote
that is not a cluster, when the protocol version supports it. This is
because P2P.Proxy does not know when it's proxying for a single node
cluster vs for a remote that is not a cluster.
This is slightly belt and suspenders, but nothing guarantees that the
peer avoids including its uuid in the SUCCESS-PLUS list as it's supposed
to. And while it probably doesn't matter if the location log is updated
redundantly, let's not find out.
This is obviously necessary in order for dropping from a cluster to be able to
drop from all nodes.
It also avoids violating numcopies when a cluster node is a special remote.
If it were used in the drop proof, nothing would prevent the cluster from
dropping from it.
I had a transfer of 3 files fail like this:
git-annex: transferrer protocol error: "(recording state in git...)"
The remote had stalldetection enabled, although I didn't see it stall.
So git-annex transferrer would have been started up. I guess that
one of these new git-annex branch reads, that happens early, caused
that message due to perhaps an uncommitted git-annex branch change.
Since the transferrer speaks a protocol over stdout, it needs to be
prevented from outputting other messages to stdout. Interestingly,
startupAnnex is run after prepRunCommand, so if a command requests quiet
output it would already be quiet. But the transferrer does not, instead
it calls Annex.setOutput SerializedOutput in its start action.
Client side support for SUCCESS-PLUS and ALREADY-HAVE-PLUS
is complete, when a PUT stores to additional repositories
than the expected on, the location log is updated with the
additional UUIDs that contain the content.
Started implementing PUT fanout to multiple remotes for clusters.
It is untested, and I fear fencepost errors in the relative
offset calculations. And it is missing proxying for the protocol
after DATA.
This assumes that the proxy for a cluster has up-to-date location
logs. If it didn't, it might proxy the checkpresent to a node that no
longer has the content, while some other node still does, and so
it would incorrectly appear that the cluster no longer contains the
content.
Since cluster UUIDs are not stored to location logs,
git-annex fsck --fast when claiming to fix a location log when
that occurred would not cause any problems. And presumably the location
tracking would later get sorted out.
At least usually, changes to the content of nodes goes via the proxy,
and it will update its location logs, so they will be accurate. However,
if there were multiple proxies to the same cluster, or nodes were
accessed directly (or via proxy to the node and not the cluster),
the proxy's location log could certainly be wrong.
(The location log access for GET has the same issues.)
Support selecting what remote to proxy for each top-level P2P protocol
message.
This only needs to be extended now to support fanout to multiple
nodes for PUT and REMOVE, and with a remote that fails for
LOCKCONTENT and UNLOCKCONTENT.
But a good first step would be to implement CHECKPRESENT and GET for
clusters. Both should select a node that actually does have the content.
That will allow a cluster to work for GET even when location tracking is
out of date.
Works down to P2P protocol.
The question now is, how to handle protocol version negotiation for
clusters? Connecting to each node to find their protocol versions and
using the lowest would be too expensive with a lot of nodes. So it seems
that the cluster needs to pick its own protocol version to use with the
client.
Then it can either negotiate that same version with the nodes when
it comes time to use them, or it can translate between multiple protocol
versions. That seems complicated. Thinking it would be ok to refuse to
use a node if it is not able to negotiate the same protocol version with
it as with the client. That will mean that sometimes need nodes to be
upgraded when upgrading the cluster's proxy. But protocol versions
rarely change.
For eg, upload fanout.
Delay connecting to a remote until it's needed. When there are many
proxied remotes, it would not do for the proxy to connect to each of
them on startup; that could take a long time.
An oversight..
And with the work in progress proxy and cluster, there
can be additional remotes that are not listed in .git/config, but are
available. Making those more discoverable is another big benefit of
this.
Handled limitCopies, as well as everything using fromNumCopies and
fromMinCopies.
This should be everything, probably.
Note that, git-annex info displays a count of repositories, which still
includes cluster. I think that's ok. It would be possible to filter out
clusters there, but to the user they're pretty much just another
repository. The numcopies displayed by eg `git-annex info .` does not
include clusters.
This is to avoid inserting a cluster uuid into the location log when
only dead nodes in the cluster contain the content of a key.
One reason why this is necessary is Remote.keyLocations, which excludes
dead repositories from the list. But there are probably many more.
Implementing this was challenging, because Logs.Location importing
Logs.Cluster which imports Logs.Trust which imports Remote.List resulted
in an import cycle through several other modules.
Resorted to making Logs.Location not import Logs.Cluster, and instead
it assumes that Annex.clusters gets populated when necessary before it's
called.
That's done in Annex.Startup, which is run by the git-annex command
(but not other commands) at early startup in initialized repos. Or,
is run after initialization.
Note that is Remote.Git, it is unable to import Annex.Startup, because
Remote.Git importing Logs.Cluster leads the the same import cycle.
So ensureInitialized is not passed annexStartup in there.
Other commands, like git-annex-shell currently don't run annexStartup
either.
So there are cases where Logs.Location will not see clusters. So it won't add
any cluster UUIDs when loading the log. That's ok, the only reason to do
that is to make display of where objects are located include clusters,
and to make commands like git-annex get --from treat keys as being located
in a cluster. git-annex-shell certainly does not do anything like that,
and I'm pretty sure Remote.Git (and callers to Remote.Git.onLocalRepo)
don't either.
Since the cluster UUID is inserted into the location log when the
location log lists a node as containing content.
Also avoid trying to lock content on cluster remotes. The cluster nodes
are also proxied, so that content can be locked on individual nodes, and
locking content on a cluster as a whole probably won't be implemented.
And made git-annex whereis use numcopies machinery for displaying its
count, so it won't count cluster UUIDs redundantly to nodes.
Other commands, like git-annex info that also display numcopies
information already used the numcopies machinery.
There is more to be done, fromNumCopies is sometimes used to get a
number that is compared with a list of UUIDs. And limitCopies doesn't
use numcopies machinery.
When imported along with Logs.Location, it can be an unused import and
it won't warn, due to reexports. The point if this is really to show
that Logs.Presence is not widely used, outside Logs/
One benefit of this is that a typo in annex-cluster-node config won't
init a new cluster.
Also it gets the cluster description set and is consistent with
initremote.
A cluster UUID is a version 8 UUID, with first octets 'a' and 'c'.
The rest of the content will be random.
This avoids a class of attack where the UUID of a repository is used as
the UUID of a cluster, which will prevent git-annex from updating
location logs for that repository. I don't know why someone would want
to do that, but let's prevent it.
Also, isClusterUUID make it easy to filter out cluster UUIDs when
writing the location logs.
