Full scope:
storagenode/{console,nodestats,notifications,reputation,storagenodedb},
web/storagenode
These columns are deprecated. They used to be for the uptime reputation
system which has been replaced by downtime tracking with audits.
Change-Id: I151d6569577d89733ac97af21a1d885323522b21
On servers with non-UTC it would have calculated a different month boundary.
If node joined in current month calculations will be related on amount of days node've been working.
Change-Id: Ie572b197f50c6cdff5a044a53dfb5b9138f82f24
Full prefix: satellite/{overlay,nodestats},storagenode/{reputation,nodestats}
Allow the storagenode to receive its audit history data from the
satellite via the satellite's GetStats endpoint.
The storagenode does not save this data for use in the API yet.
Change-Id: I9488f4d7a4ccb4ccf8336b8e4aeb3e5beee54979
Add the upload size to the log lines of the storagenode Upload endpoint
to provides the information to Storage node operators.
Change-Id: Ife661d28be72c2bf02579093e21fa811566ac8dd
We want to stop using the serial_numbers table in satelliteDB. One of the last places using the serial_numbers table is when storagenodes settle orders, we look up the bucket name and project ID from the serial number from the serial_numbers table.
Now that we have support to add encrypted metadata into the OrderLimit, this PR makes use of that and now attempts to read the project ID and bucket name from the encrypted orderLimit metadata instead of from the serial_numbers table. For backwards compatibility and to ensure no errors, we will still fallback to the old way of getting that info from the serial_numbers table, but this will be removed in the next release as long as there are no errors.
All processes that create orderLimits must have an orders.encryption-keys set. The services that create orderLimits (and thus need to encrypt the order metadata) are the satellite apiProcess, the repair process, audit service (core process), and graceful exit (core process). Only the satellite api process decrypts the order metadata when storagenodes settle orders. This means that the same encryption key needs to be provided in the config for the satellite api process, repair process, and the core process like so:
orders.include-encrypted-metadata=true
orders.encryption-keys="<"encryptionKeyID>=<encryptionKey>"
Change-Id: Ie2c037971713d6fbf69d697bfad7f8b672eedd66
Before manipulating order information on storagenodes we need to wait
for the orders to propagate to the database. Some of that happens
async with uplink.
Change-Id: Iaacfd7db0909ab5d2831d06388e5fb27b6d4778f
Define constants of 32 KiB as the upper limit of the marshalled order
and limit protobuf sizes. This value gives lots of buffer in case the
protobufs ever change, but is not as extreme as what we were doing
before in V0 files, which was to use the Uint32 max value.
Change-Id: I0914d17dde3b044b2611af33f931d46d55f81e98
Fix the error message reported by a wrong order size due to passing the
wrong variable to the interpolation pattern.
Change-Id: Ic0059615c60cfa33a26d4aeb0ebda5e586f0df05
`make` built function to build a new slice with a negative
length panics.
`make` length parameter is of `int` type.
These changes avoid that `make` panics on 32 bits architecture due to
the fact that `int` type is a `int32` an uint32 value can be over the
maximum `int32`, and when that happens the length parameter value
becomes negative and makes `make` to panic.
Change-Id: Ife9ab5993916d6dcf5584b37c208272269cb2b45
If the satellite fails to pingback the storage node during CheckIn
an error message is returned to the node in the response, but the actual
error value returned is nil. We are only checking the error. This means
the node has no feedback about the failure, and the node also does not
attempt to retry the connection.
Change-Id: Iaed00e422ba91af573e72255cc6671ea97928eae
This change fixes two things which can make reading from a corrupted
orders file inefficient.
* When a corrupted order is detected, but the underlying error is an
UnexpectedEOF (as opposed to a pb.Unmarshal error, for instance), there
is no point in attempting to read from the file another time to find an
additional uncorrupted order - we will continue to get UnexpectedEOF
errors until we seek to the very end of the file and get a normal EOF.
Instead, when UnexpectedEOF occurs, log and send metrics as with other
types of corruption, but do not attempt to read again.
* When a corrupted order is detected, instead of seeking forward only
one byte for the next attempt, seek forward by the size of entryHeader.
This cuts down on the number of iterations needed to find an uncorrupted
order after detecting a corrupted one.
Change-Id: Ie1a613127e29d29318584ec7f60e8f7554f73487
Previously, we created a new file to use for directory verification
every time the storage node starts. This is not helpful if the storage node
points to the wrong directory when restarting. Now we will only create the file
on setup. Now the file should be created only once and will be verified at
runtime.
Change-Id: Id529f681469138d368e5ea3c63159befe62b1a5b
We shouldn't have any EOF issues with recent drpc fix, let's reenable
and see whether it's still flaky.
Change-Id: I0de312bcb087c7f70ec9d3281d73d86f971845d5
Make metainfo.RSConfig a valid pflag config value. This allows us to
configure the RSConfig as a string like k/m/o/n-shareSize, which makes
having multiple supported RS schemes easier in the future.
RS-related config values that are no longer needed have been removed
(MinTotalThreshold, MaxTotalThreshold, MaxBufferMem, Verify).
Change-Id: I0178ae467dcf4375c504e7202f31443d627c15e1
Missed one case of Unmarshal in the previous commit for V0 files (0f4e4969b7)
In V1, unmarshalling was being attempted before the checksum was
verified, so this commit moves those calls to the end of the V1 ReadOne
function.
Change-Id: Ic0b49f0bbc91fb61fb28af6003060994d0af22ed
In V0 orders files, unexpected EOF is correctly treated as a file
corruption, but pb.Unmarshal can also fail, and this is not treated as a
file corruption. This commit fixes that.
Change-Id: I6b446a10f4b1a5a44e832cbcc9bf8b2548cfcfeb
In production we are seeing ~115 storage nodes (out of ~6,500) are not using the new SettlementWithWindow endpoint (but they are upgraded to > v1.12).
We analyzed data being reported by monkit for the nodes who were above version 1.11 but were not successfully submitting orders to the new endpoint.
The nodes fell into a few categories:
1. Always fail to list orders from the db; never get to try sending orders from the filestore
2. Successfully list/send orders from the db; never get to calling satellite endpoint for submitting filestore orders
3. Successfully list/send orders from the db; successfully list filestore orders, but satellite endpoint fails (with "unauthenticated" drpc error)
The code change here add the following to address these issues:
- modify the query for ordersDB.listUnsentBySatellite so that we no longer select expired orders from the unsent_orders table
- always process any orders that are in the ordersDB and also any orders stored in the filestore
- add monkit monitoring to filestore.ListUnsentBySatellite so that we can see the failures/successes
Change-Id: I0b473e5d75252e7ab5fa6b5c204ed260ab5094ec
WHAT:
changed estimation table row order.
WHY:
to show gross total for selected period to avoid misunderstanding
when held amount is bigger than paid multiple times.
Change-Id: I03881c8af682372139a378030acf04f199d3260b
Use tagsql.DB pointer as step database, to propagate changes
back and forth between actual database and migration.
Adds CreateDB operation to the migration step to be able to
create new dbs before executing migration action.
Adjusts storagenode database migration to use inner tagsql.DB
pointer of each database as step.DB.
Adjusts satellite dabase migration, adds proxy migrationDB field
to satellite db that wraps itself as tagsql.DB, pointer of which
is used as step.DB.
Change-Id: Ifed4de5b01a356cf7b37db64d2eaeb7b61982c5c
Right now, the best way for a storage node operator to get the current
space used for each satellite is to run the `storagenode exit-satellite`
command for graceful exit, and cancel at the second confirmation prompt.
This is convoluted and the data is readily available from the Blobs
Usage Cache.
This change adds the current space used by each satellite to the
endpoints `/api/sno` and `/api/sno/satellite/<Satellite ID>`
Change-Id: I2173005bb016fc76db96fd598d26b485e5b2aa0b
V1 allows the storagenode to continue reading orders from an
unsent/archived orders file, even if orders in the middle are corrupted.
Change-Id: Iea4117d55c05ceeb77f47d5c973e5ba95da46c66
log.Fatal immediately terminates the program without running any defers.
We should properly close all the services and databases.
Change-Id: I5e959cef3eafedeacb3a2062e3da47e8d04e8e75
Accidentally we imported the wrong monkit package with a previous
commit and made our go.mod and go.sum file unclean.
This should fix it.
Change-Id: I4c3c8b696f59cfd06dc2d5436bb7aea2805936ce
Since storage nodes check to see if any order files can be sent every 5
minutes, every storage node attempts to send orders to the satellite
within 5 minutes of each hour since this is when the files become
"available" to send. It is placing a lot of load on our satellite and
storage nodes are not being paid out properly due to timeouts during
order sending due to the increased satellite load.
Change-Id: I44d991b5884b8c11e8a3856d39aee8323f086b51
Abstract details of writing and reading data to/from orders files so
that adding V1 and future maintenance are easier.
Change-Id: I85f4a91761293de1a782e197bc9e09db228933c9
Combine store.writeLimit and store.writeOrder into
store.writeLimitAndOrder, which only requires a single call to
file.Write(). This simplifies code, but it also reduces the likelihood
of multiple calls to Write() increasing the likelihood of file
corruption.
Also combine the corresponding readLimit/readOrder functions for
consistency.
Change-Id: I62ed406fa2c02708465a678d18293f510f666440
We only need to lock aquire mutexes inside ListUnsentBySatellite when we
want to determine whether a file has an active enqueue in progress.
On some nodes, ListUnsentBySatellite can take a particularly long time, having
undesired side-effects, so if we can minimize locking time, those nodes
will be better off.
Also, lock archive mu during ListUnsentBySatellite so files cannot be
archived and listed at the same time.
Change-Id: Ieb7e2a759c20c724a74dd8315728c873ccab14a3
If we see an UnexpectedEOF error when attempting to read orders, return
the orders we have been able to read successfully and do not return an
error. This behavior ensures that the storagenode orders service
attempts to archive corrupted files and does not retry them repeatedly
and get stuck.
Change-Id: I0d00d1e174f968af6e99ca861eddad190f1339e2
Additionally, this PR changes NewNodeFraction devDefault and testplanet config from 0.05 to 1.
This is because many tests relied on selecting nodes that were reputable based on audit and uptime
counts of 0, in effect, selecting new nodes as reputable ones.
However, since reputation is now indicated by a vetted_at db field that is explicitly set
rather than implied by audit and uptime counts, it would be more complicated to try to
update all of the nodes' reputations before selecting nodes for tests.
Now we just allow all test nodes to be new if needed.
Change-Id: Ib9531be77408662315b948fd029cee925ed2ca1d
periodically create and delete a temp file in the storage directory
to verify writability. If this check fails, shut the node down.
Change-Id: I433e3a8d1d775fc779ae78e7cf3144a05ffd0574
satellite.DB.Console().Projects().GetAll database query
can be replaced with planet.Uplinks[0].Projects[0].ID
Change-Id: I73b82b91afb2dde7b690917345b798f9d81f6831
When we call ordersStore.BeginEnqueue, the unsent orders file for that
satellite and hour is prevented from being sent. It is freed when the
commit callback returned by BeginEnqueue is used. This change ensures
that we always call the commit callback, even when we have an empty
order or an order with Amount <= 0.
Change-Id: Ic4678f7eaa1e6957dd77d4bb5a23bb35d25b1e93
This change accomplishes multiple things:
1. Instead of having a max in flight time, which means
we effectively have a minimum bandwidth for uploads
and downloads, we keep track of what windows have
active requests happening in them.
2. We don't double check when we save the order to see if it
is too old: by then, it's too late. A malicious uplink
could just submit orders outside of the grace window and
receive all the data, but the node would just not commit
it, so the uplink gets free traffic. Because the endpoints
also check for the order being too old, this would be a
very tight race that depends on knowledge of the node system
clock, but best to not have the race exist. Instead, we piggy
back off of the in flight tracking and do the check when
we start to handle the order, and commit at the end.
3. Change the functions that send orders and list unsent
orders to accept a time at which that operation is
happening. This way, in tests, we can pretend we're
listing or sending far into the future after the windows
are available to send, rather than exposing test functions
to modify internal state about the grace period to get
the desired effect. This brings tests closer to actual
usage in production.
4. Change the calculation for if an order is allowed to be
enqueued due to the grace period to just look at the
order creation time, rather than some computation involving
the window it will be in. In this way, you can easily
answer the question of "will this order be accepted?" by
asking "is it older than X?" where X is the grace period.
5. Increases the frequency we check to send up orders to once
every 5 minutes instead of once every hour because we already
have hour-long buffering due to the windows. This decreases
the maximum latency that an order will be reported back to
the satellite by 55 minutes.
Change-Id: Ie08b90d139d45ee89b82347e191a2f8db1b88036
On run, write the storage directory verification file.
Every time the node runs it will write the file even if it already exists.
The reason we do this is because if the verification file is missing, the SN
doesn't know whether it is an incorrect directory, or it simply hasn't written
the file yet, and we want to keep nodes running without needing operator intervention.
Once this change has been a part of the minimum version for several releases,
we will move the file creation from the run command to the setup
command. Run will only verify its existence.
Change-Id: Ib7d20e78e711c63817db0ab3036a50af0e8f49cb
Sometimes SNOs fail to properly configure or lose connection to their storage directory
which can result in DQ. This causes unnecessary repair and is unfortunate for all parties.
This change introduces the creation of a special file in the storage directory at runtime
containing the node ID. While the storage node runs, it periodically verifies that it can
find said file with the correct contents in the correct location. If not, the node will
shut down with an error message.
This change will solve the issue of nodes losing access to the storage directory, but it will not
solve the issue of nodes pointing to the wrong directory, as the identifying file is created each
time the node starts up. After this change has been the minimum version for a few releases, we will
remove the creation of the directory-identifying file from the storage node run command and add it
to the setup command.
Change-Id: Ib7b10e96ac07373219835e39239e93957e7667a4
* Add all new orders to the orders filestore instead of the database.
* Submit orders from the filestore to the new satellite SettleWindow
endpoint.
The orders filestore will eventually replace the orders DB completely.
For now, we will still be checking the orders DB and submitting those
orders if they exist. In a later release, we will completely remove the
orders DB, but we need both the DB and filestore for the transitionary
period.
Change-Id: Iac8780fd5ab770296181bbd313e1d335f072d4dc
A failed test showed the same piece being deleted twice. This happens if the graceful exit completes before a previous piece deletion finishes. This change adds a "wait" on the limiter before executing the delete all step when GE is done.
Change-Id: I1c8c49d1e501c2728c80d4224a4854e742be27da
To prevent storagenode from implicitly recreating missing dbs and storage,
as such behaviour leads to audit failures. Do not allow storagenode to
start if any of dbs or storage is missing, corrupted, or dedicated storage disk is
unmounted, to get downtime instead.
Change-Id: Ic64e1f0ff4d8ef5b2fddbe7a7e53df4f4bd8652e
We only sync free disk and available space, if necessary, on startup. If the
SNs disk fills up with non-storj data, we will not know about it when reporting
available space to the satellite.
Solution: whenever we check the node's capacity, double check free disk.
If free disk < than expected available space, return free disk.
Change-Id: I66265c16e03be45b6e1f5817c70df7eac0a76455
* Instead of archiving a list of orders and deleting an "unsent" file in
separate steps, archival simply moves the old unsent file to a new
archived file
* Add maxInFlightTime to be used along with grace period for sending
buffer
* Create unsent/archival directories in constructor
* Code cleanup
Change-Id: Ia3bc2aaf60cced6c6d413465423d78c7d5151188
Nodes are receiving an error that heldamount rpc doesn't exist on a
satellite. This simply adds which satellite to the error.
Change-Id: I7708e0511b55fdd2425969db2a545645339bad81
on node's start checks if any of trusted satellites has GE status "Exited successfully"
if so - trying to delete blobs/satellite folder, so no trash left on SNO.
Change-Id: I566266c84f2a872df54cd01bc2f15a9934f138ed
Per https://documentation.tardigrade.io/pricing/billing-and-payment:
"The calculation of per object fees is based on a standard 720-hour month."
On most years, the average value is 730 (365*24/12), except leap years.
However, we want to have ours be 720 (30*24) so its lines up with days.
Change-Id: Ifb9691878f1a7ea81ed36c92b37985493295fe31
When an uplink or repair work finishes uploading a piece to a
storagenode, it has no reason to wait another round trip after the piece
is committed to gracefully close the connection - in many (most?) cases,
the connection is simply canceled once the upload is complete. This has
the unintended side effect of producing a lot of "piece canceled" logs
and metrics on the storagenode side, when the reality is that the piece
uploads were successful, and not really canceled. This commit fixes
that.
Change-Id: Icbc1f7857d380134560219c1c19c186df2783cd0
with 128B/sec, a satellite with 10min default timeout could already
closed its connection to a node even though the node was able to
compelete the transfer.
Change-Id: I6173d6473a62c6d0b0e0a8765c1ae0a5e57b0a08
* Allow orders to be archived after being settled successfully with the
satellite.
* Allow for cleanup of orders that were archived before a certain time.
* Rewrite other parts of the orders file store to work better with new
design.
Change-Id: I39bea96d80e66a324ec522745169bd6d8b351751
This will give storagenode operators a better idea of whether the memory
allocated to the usedserials store is sufficient.
Change-Id: I5c30f2e39473a573f43409511ad9e2e32680479c
* Will replace order limits database.
* This change adds functionality for storing and listing unsent orders.
* The next change will add functionality for order archival after
submission.
Change-Id: Ic5e2abc63991513245b6851a968ff2f2e18ce48d
Part 2 of moving usedserials in memory
* Drop usedserials table in storagenodedb
* Use in-memory usedserials store in place of db for order limit
verification
* Update order limit grace period to be only one hour - this means
uplinks must send their order limits to storagenodes within an hour of
receiving them
Change-Id: I37a0e1d2ca6cb80854a3ef495af2d1d1f92e9f03
Implement an in-memory store for keeping track of order limit serial
numbers. It automatically deletes items if its size exceeds a configured
limit.
This change is part 1 - it creates the store
In part 2, the in-memory store will replace the usedserials database
Change-Id: I36f540ed809f034a27c1d7cede8a0a8b080af818
..before they are transferred to another node and submitted to the
satellite as successful piece transfers, because if we submit an invalid
signature, the node will be marked as a cheater and disqualified
immediately.
These signatures should have been validated when the piece was
originally stored, but bitrot does happen and needn't be cause for an
immediate DQ.
Change-Id: I8b0ebd5812ea8a2e60766005b7251fbb74ef7857
In walkNamespaceWithPrefix log in case of "lstat" error, because this may indicate an underlying disk corruption.
SG-50
Change-Id: I867c3ffc47cfac325ae90658ec4780d213ff3e63
Most places now need the NodeURL rather than the ID and Address
separately. This simplifies code in multiple places.
Change-Id: I52621d8ca52296a8b5bf7afbc1001cf8bfb44239
See https://storjlabs.atlassian.net/browse/SM-752
These changes allow us to change the log level at runtime through a handler off of the debug endpoint.
Examples of changing the log level on storj-sim
To get the current level for the satellite api process:
curl -XGET 'http://127.0.0.1:10009/logging' --header 'Content-Type: text/plain'
To change the log level:
curl -XPUT 'http://127.0.0.1:10009/logging' --header 'Content-Type: text/plain' --data-raw '{"level":"error"}'
Change-Id: I05d164b290929fa06b6d78c01075ee41f8238044
This test was the last place using it. Replace it with a direct call so
we can remove the method from uplink piecestore.
Change-Id: I62e13028663a7e67aa2495f90ecc02d0d8657fbd
Currently uploads can cause a lot of IOPS, reduce this by introducing a
in-memory buffer on-top of the file.
Change-Id: I5f4e3e01c0a36258271d180b922107de447bcb59
it was being used in ways that implied it should be NOT NULL
even though it was possibly null. we used to get this data
from the satellite db's added_at column as seen in 30369b02,
so backfill it using that data where joined_at is NULL, and
then alter the table to constrain the column to be NOT NULL.
Fixes#3866.
Change-Id: If2d856189209740d985f71dada7b93525e625ef3
According to the docs at https://www.sqlite.org/lang_altertable.html
doing the steps
1. Rename old table
2. Create new table
3. Copy data
4. Drop old table
is incorrect and should be
1. Create new table
2. Copy data
3. Drop old table
4. Rename new into old
Additionally, each step was being run in a different transaction,
which could cause permanent failures if a problem happened during
the migration.
Avoid both of those problems by changing up some previous migrations
that ran in this way. Since they are semantically identical, it's
fine to change up these old migrations. It will help make newer
nodes coming up for the first time more robust.
Change-Id: I43fb004fa1b6cb2fe2554f9920925420da28fb4a
CreateTables hasn't been quite true for a while now, rename to
MigrateToLatest to be clearer in it's behavior.
Change-Id: Ida48e95122a5d9b7a814e922d3698e00024a2ba7
Before the deleter would close its done channel once, so if additional
tests shared a storagenode, even if not in parallel, the later waits
would not work properly. This fixes that problem.
Change-Id: I7dcacf6699cef7c2c2948ba0f4369ef520601bf5
Currently this test was the last place that was using
piecestore.Client.DeletePieces. This way we can remove it from uplink to
reduce the code.
Change-Id: I72fda8888d05181f95eeb544d067c031ec3e36a0
Currently Cockroach isn't performant for concurrent database setup and
tear-down. Instead of a single instance allow setting multiple potential
connection strings and let the tests pick one connection string
randomly.
This improves test duration by ~10 minutes.
While we are at significantly changing how pgtest works, introduce
helper PickPostgres and PickCockroach for selecting the database to
reduce code duplications in multiple places.
Change-Id: I8ad171d5c4c8a4fc081ec2ae9bdd0cc948a80619
There was a race in the test code for piece deleter, which made it
possible to broadcast on the condition variable before anyone was
waiting. This change fixes that and has Wait take a context so it times
out with the context.
Change-Id: Ia4f77a7b7d2287d5ab1d7ba541caeb1ba036dba3
When we receive a piece deletion request, include the number of piece
IDs we couldn't add to the queue in the reponse
Change-Id: Ibebbe92ac50105bb5c74b18211ed38d468eb33f3
Each time we process a piece deletion on the storagenode, monitor how
long the item was in the queue and the size of the queue.
Change-Id: I23f1a44f8b9cecb901bdf4739d55c005ffed4bef
To improve delete performance, we want to process deletes asynchronously
once the message has been received from the satellite. This change makes
it so that storagenodes will send the delete request to a piece Deleter,
which will process a "best-effort" delete asynchronously and return a
success message to the satellite.
There is a configurable number of max delete workers and a max delete
queue size.
Change-Id: I016b68031f9065a9b09224f161b6783e18cf21e5
