In the repair subsystem, it is necessary to acquire several extra
properties of nodes that are holding pieces of things or may be
selected to hold pieces. We need to know if a node is 'online' (the
definition of "online" may change somewhat depending on the situation),
if a node is in the process of graceful exit, and whether a node is
suspended. We can't just filter out nodes with all of these properties,
because sometimes we need to know properties about nodes even when the
nodes are suspended or gracefully exiting.
I thought the best way to do this was to add fields to SelectedNode,
and (to avoid any confusion) arrange for the added fields to be
populated wherever SelectedNode is returned, whether or not the new
fields are necessarily going to be used.
If people would rather I use a separate type from SelectedNode, I can do
that instead.
Change-Id: I7804a0e0a15cfe34c8ff47a227175ea5862a4ebc
Current node selection logic (in case of using SelectBySubnet):
1. selects one subnet randomly
2. selects one node randomly from the subnet
3. applies the placement NodeFilters to the node and ignore it, if doesn't match
This logic is wrong:
1. Imagine that we have a subnet with two DE and one GB nodes.
2. We would like to select DE nodes
2. In case of GB node is selected (randomly) in step2, step3 will ignore the subnet, even if there are good (DE) nodes in there.
Change-Id: I7673f52c89b46e0cc7b20a9b74137dc689d6c17e
We don't need it any more, as CountryCode uses location.Set, which supports exclusion (`Without`).
Change-Id: Ie311ae19fefa0bc9a0161496af1233ef4a6607df
stats/size/count is not used by any production code, and it's not required, as we can assert the state with other checks.
real motivation: next commits will make the Selector of the State configurable, therefore we won't have one single Stat, it depends on the request parameters.
(we plan to support both network and id based randomization)
Change-Id: I631828fc0046d2fef5b7a674fc0268a0446e9655
* Handle failed country code conversion.
* Avoid potential issues with a data-race due to shared slice.
Updates #6028
Change-Id: If7beef2619abd084e1f4109de2d323f834a6090a
All the files in uploadselection are (in fact) related to generic node selection, and used not only for upload,
but for download, repair, etc...
Change-Id: Ie4098318a6f8f0bbf672d432761e87047d3762ab
We use two different Node types in `overlay` and `uploadnodeselection` and converting back and forth.
Using the same object would allow us to use a unified node selection interface everywhere.
Change-Id: Ie71e29d60184ee0e5b4547eb54325f09c418f73c
We were using the UploadSelectionCache previously, which does _not_ have
all nodes, or even all online nodes, in it. So all nodes with less than
MinimumVersion, or with less than MinimumDiskSpace, or nodes suspended
for unknown audit errors, or nodes that have started graceful exit, were
all missing, and ended up having empty last_nets. Even with all that,
I'm kind of surprised how many nodes this involved, but using the upload
selection cache was definitely wrong.
This change uses the download selection cache instead, which excludes
nodes only when they are disqualified, gracefully exited (completely),
or offline.
Change-Id: Iaa07c988aa29c1eb05796ac48a6f19d69f5826c1
The query for GetNodesNetworkInOrder is causing far too much load on the
database. Since it is not critical that the repair checker have
perfectly up-to-date node network information, we can use a cache
instead.
Change-Id: I07ad45bfdeb46529da093941a06c2da8a00ce878
Up to now, we have been implementing the DistinctIP preference with code
in two places:
1. On check-in, the last_net is determined by taking the /24 or /64
(in ResolveIPAndNetwork()) and we store it with the node record.
2. On node selection, a preference parameter defines whether to return
results that are distinct on last_net.
It can be observed that we have never yet had the need to switch from
DistinctIP to !DistinctIP, or from !DistinctIP to DistinctIP, on the
same satellite, and we will probably never need to do so in an automated
way. It can also be observed that this arrangement makes tests more
complicated, because we often have to arrange for test nodes to have IP
addresses in different /24 networks (a particular pain on macOS).
Those two considerations, plus some pending work on the repair framework
that will make repair take last_net into consideration, motivate this
change.
With this change, in the #2 place, we will _always_ return results that
are distinct on last_net. We implement the DistinctIP preference, then,
by making the #1 place (ResolveIPAndNetwork()) more flexible. When
DistinctIP is enabled, last_net will be calculated as it was before. But
when DistinctIP is _off_, last_net can be the same as address (IP and
port). That will effectively implement !DistinctIP because every
record will have a distinct last_net already.
As a side effect, this flexibility will allow us to change the rules
about last_net construction arbitrarily. We can do tests where last_net
is set to the source IP, or to a /30 prefix, or a /16 prefix, etc., and
be able to exercise the production logic without requiring a virtual
network bridge.
This change should be safe to make without any migration code, because
all known production satellite deployments use DistinctIP, and the
associated last_net values will not change for them. They will only
change for satellites with !DistinctIP, which are mostly test
deployments that can be recreated trivially. For those satellites which
are both permanent and !DistinctIP, node selection will suddenly start
acting as though DistinctIP is enabled, until the operator runs a single
SQL update "UPDATE nodes SET last_net = last_ip_port". That can be done
either before or after deploying software with this change.
I also assert that this will not hurt performance for production
deployments. It's true that adding the distinct requirement to node
selection makes things a little slower, but the distinct requirement is
already present for all production deployments, and they will see no
change.
Refs: https://github.com/storj/storj/issues/5391
Change-Id: I0e7e92498c3da768df5b4d5fb213dcd2d4862924
For nodes in excluded areas, we don't necessarily want to remove them
from the pointer, but we do want to increase the number of pieces in the
segment in case those excluded area nodes go down. To do that, we
increase the number of pieces repaired by the number of pieces in
excluded areas.
Change-Id: I0424f1bcd7e93f33eb3eeeec79dbada3b3ea1f3a
Create global config to specify a list of country codes that should be
excluded from node selection during uploads.
This exclusion is not implemented when the upload selection cache is
disabled.
Change-Id: Ic41e8b4f18857a11045668eac23107da99668a72
This commit doesn't change any behavior, just organize the code in
different way to make it easier to implement different Criterias
to include nodes. Today we use NodeID and Subnet based selection
but later Criteria can be extended with different kind of
placement rules (like geofencing).
The change nodeselection is used by segment allocaton (upload) and repair
and excludes nodes from an in-memory selection.
Resolves https://github.com/storj/storj/issues/4240
Change-Id: I0c1955fe16a045e3b76d7e50b2e1f4575a7ff095
Currently nodeselection package only contained state for uploads, move
these to a subpackage, such that we can make another "downloadselection"
for downloads. Then move selection logic from overlay to nodeselection.
Change-Id: I0fc42bcae3a29db2728dae9f3863b1e95bf5165b
* Deduplicate NodeID list prior to fetching IPs.
* Use NodeSelectionCache for fetching reliable IPs.
* Return number of segements, reliable pieces and all pieces.
Change-Id: I13e679caab275488b4037624b840a4068dad9589
Currently node selection cache is biased towards the same subnet. This
implements static node selection for distinct such that it selects with
equal probability subnets rather than id-s.
This is mostly a copy paste + modifications from previous node selection
state.
Change-Id: Ia5c0aaf68e7feca78fbbd7352ad369fcb77c3a05
