Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

fix(OntologyResponderV2): Add a global ontology cache lock #1637

Merged
merged 4 commits into from Apr 27, 2020
Merged

fix(OntologyResponderV2): Add a global ontology cache lock #1637

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
a4b9f18
fix(OntologyResponderV2): Add a global ontology cache lock.
Apr 22, 2020
f8a2586
fix: Increase IriLocker max retries.
Apr 22, 2020
8f4aff5
Merge branch 'develop' into fix/DSP-181-ontology-updates
Apr 24, 2020
541267d
test(IriLocker): Fix test timeouts.
Apr 24, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
2 changes: 1 addition & 1 deletion webapi/src/main/scala/org/knora/webapi/responders/IriLocker.scala
View file
Open in desktop
Expand Up @@ -60,7 +60,7 @@ object IriLocker {
/**
* The number of times to try to acquire a lock before giving up.
*/
private val MAX_LOCK_RETRIES = 40
private val MAX_LOCK_RETRIES = 150

/**
* The total number of milliseconds that an API request should wait before giving up on acquiring a lock.
Expand Down
61 changes: 31 additions & 30 deletions webapi/src/main/scala/org/knora/webapi/responders/v2/OntologyResponderV2.scala
View file
Open in desktop
Expand Up @@ -61,16 +61,17 @@ import scala.concurrent.Future
*/
class OntologyResponderV2(responderData: ResponderData) extends Responder(responderData) {

/**
* The name of the ontology cache.
*/
// The name of the ontology cache.
private val OntologyCacheName = "ontologyCache"

/**
* The cache key under which cached ontology data is stored.
*/
// The cache key under which cached ontology data is stored.
private val OntologyCacheKey = "ontologyCacheData"

// The global ontology cache lock. This is needed because every ontology update replaces the whole ontology cache
// (because definitions in one ontology can refer to definitions in another ontology). Without a global lock,
// concurrent updates (even to different ontologies) could overwrite each other.
private val ONTOLOGY_CACHE_LOCK_IRI = "http://rdfh.ch/ontologies"

/**
* The in-memory cache of ontologies.
*
Expand All @@ -97,7 +98,7 @@ class OntologyResponderV2(responderData: ResponderData) extends Responder(respon
case StandoffEntityInfoGetRequestV2(standoffClassIris, standoffPropertyIris, requestingUser) => getStandoffEntityInfoResponseV2(standoffClassIris, standoffPropertyIris, requestingUser)
case StandoffClassesWithDataTypeGetRequestV2(requestingUser) => getStandoffStandoffClassesWithDataTypeV2(requestingUser)
case StandoffAllPropertyEntitiesGetRequestV2(requestingUser) => getAllStandoffPropertyEntitiesV2(requestingUser)
case CheckSubClassRequestV2(subClassIri, superClassIri, requestingUser) => checkSubClassV2(subClassIri, superClassIri, requestingUser)
case CheckSubClassRequestV2(subClassIri, superClassIri, requestingUser) => checkSubClassV2(subClassIri, superClassIri)
case SubClassesGetRequestV2(resourceClassIri, requestingUser) => getSubClassesV2(resourceClassIri, requestingUser)
case OntologyKnoraEntityIrisGetRequestV2(namedGraphIri, requestingUser) => getKnoraEntityIrisInNamedGraphV2(namedGraphIri, requestingUser)
case OntologyEntitiesGetRequestV2(ontologyIri, allLanguages, requestingUser) => getOntologyEntitiesV2(ontologyIri, allLanguages, requestingUser)
Expand Down Expand Up @@ -1365,7 +1366,7 @@ class OntologyResponderV2(responderData: ResponderData) extends Responder(respon
* @param superClassIri the IRI of the resource or value class to check for (whether it is a a super class of `subClassIri` or not).
* @return a [[CheckSubClassResponseV2]].
*/
private def checkSubClassV2(subClassIri: SmartIri, superClassIri: SmartIri, requestingUser: UserADM): Future[CheckSubClassResponseV2] = {
private def checkSubClassV2(subClassIri: SmartIri, superClassIri: SmartIri): Future[CheckSubClassResponseV2] = {
for {
cacheData <- getCacheData
response = CheckSubClassResponseV2(
Expand Down Expand Up @@ -1771,10 +1772,10 @@ class OntologyResponderV2(responderData: ResponderData) extends Responder(respon
// Make the internal ontology IRI.
internalOntologyIri = stringFormatter.makeProjectSpecificInternalOntologyIri(validOntologyName, createOntologyRequest.isShared, projectInfo.project.shortcode)

// Do the remaining pre-update checks and the update while holding an update lock on the ontology.
// Do the remaining pre-update checks and the update while holding a global ontology cache lock.
taskResult <- IriLocker.runWithIriLock(
apiRequestID = createOntologyRequest.apiRequestID,
iri = internalOntologyIri.toString,
iri = ONTOLOGY_CACHE_LOCK_IRI,
task = () => makeTaskFuture(internalOntologyIri)
)
} yield taskResult
Expand Down Expand Up @@ -1848,10 +1849,10 @@ class OntologyResponderV2(responderData: ResponderData) extends Responder(respon
_ <- checkExternalOntologyIriForUpdate(changeOntologyMetadataRequest.ontologyIri)
internalOntologyIri = changeOntologyMetadataRequest.ontologyIri.toOntologySchema(InternalSchema)

// Do the remaining pre-update checks and the update while holding an update lock on the ontology.
// Do the remaining pre-update checks and the update while holding a global ontology cache lock.
taskResult <- IriLocker.runWithIriLock(
apiRequestID = changeOntologyMetadataRequest.apiRequestID,
iri = internalOntologyIri.toString,
iri = ONTOLOGY_CACHE_LOCK_IRI,
task = () => makeTaskFuture(internalOntologyIri = internalOntologyIri)
)
} yield taskResult
Expand Down Expand Up @@ -2027,10 +2028,10 @@ class OntologyResponderV2(responderData: ResponderData) extends Responder(respon
internalClassIri = externalClassIri.toOntologySchema(InternalSchema)
internalOntologyIri = externalOntologyIri.toOntologySchema(InternalSchema)

// Do the remaining pre-update checks and the update while holding an update lock on the ontology.
// Do the remaining pre-update checks and the update while holding a global ontology cache lock.
taskResult <- IriLocker.runWithIriLock(
apiRequestID = createClassRequest.apiRequestID,
iri = internalOntologyIri.toString,
iri = ONTOLOGY_CACHE_LOCK_IRI,
task = () => makeTaskFuture(
internalClassIri = internalClassIri,
internalOntologyIri = internalOntologyIri
Expand Down Expand Up @@ -2387,10 +2388,10 @@ class OntologyResponderV2(responderData: ResponderData) extends Responder(respon
internalClassIri = externalClassIri.toOntologySchema(InternalSchema)
internalOntologyIri = externalOntologyIri.toOntologySchema(InternalSchema)

// Do the remaining pre-update checks and the update while holding an update lock on the ontology.
// Do the remaining pre-update checks and the update while holding a global ontology cache lock.
taskResult <- IriLocker.runWithIriLock(
apiRequestID = addCardinalitiesRequest.apiRequestID,
iri = internalOntologyIri.toString,
iri = ONTOLOGY_CACHE_LOCK_IRI,
task = () => makeTaskFuture(
internalClassIri = internalClassIri,
internalOntologyIri = internalOntologyIri
Expand Down Expand Up @@ -2552,10 +2553,10 @@ class OntologyResponderV2(responderData: ResponderData) extends Responder(respon
internalClassIri = externalClassIri.toOntologySchema(InternalSchema)
internalOntologyIri = externalOntologyIri.toOntologySchema(InternalSchema)

// Do the remaining pre-update checks and the update while holding an update lock on the ontology.
// Do the remaining pre-update checks and the update while holding a global ontology cache lock.
taskResult <- IriLocker.runWithIriLock(
apiRequestID = changeCardinalitiesRequest.apiRequestID,
iri = internalOntologyIri.toString,
iri = ONTOLOGY_CACHE_LOCK_IRI,
task = () => makeTaskFuture(
internalClassIri = internalClassIri,
internalOntologyIri = internalOntologyIri
Expand Down Expand Up @@ -2653,10 +2654,10 @@ class OntologyResponderV2(responderData: ResponderData) extends Responder(respon
internalClassIri = externalClassIri.toOntologySchema(InternalSchema)
internalOntologyIri = externalOntologyIri.toOntologySchema(InternalSchema)

// Do the remaining pre-update checks and the update while holding an update lock on the ontology.
// Do the remaining pre-update checks and the update while holding a global ontology cache lock.
taskResult <- IriLocker.runWithIriLock(
apiRequestID = deleteClassRequest.apiRequestID,
iri = internalOntologyIri.toString,
iri = ONTOLOGY_CACHE_LOCK_IRI,
task = () => makeTaskFuture(
internalClassIri = internalClassIri,
internalOntologyIri = internalOntologyIri
Expand Down Expand Up @@ -2771,10 +2772,10 @@ class OntologyResponderV2(responderData: ResponderData) extends Responder(respon
internalPropertyIri = externalPropertyIri.toOntologySchema(InternalSchema)
internalOntologyIri = externalOntologyIri.toOntologySchema(InternalSchema)

// Do the remaining pre-update checks and the update while holding an update lock on the ontology.
// Do the remaining pre-update checks and the update while holding a global ontology cache lock.
taskResult <- IriLocker.runWithIriLock(
apiRequestID = deletePropertyRequest.apiRequestID,
iri = internalOntologyIri.toString,
iri = ONTOLOGY_CACHE_LOCK_IRI,
task = () => makeTaskFuture(
internalPropertyIri = internalPropertyIri,
internalOntologyIri = internalOntologyIri
Expand Down Expand Up @@ -2872,10 +2873,10 @@ class OntologyResponderV2(responderData: ResponderData) extends Responder(respon
_ <- checkExternalOntologyIriForUpdate(deleteOntologyRequest.ontologyIri)
internalOntologyIri = deleteOntologyRequest.ontologyIri.toOntologySchema(InternalSchema)

// Do the remaining pre-update checks and the update while holding an update lock on the ontology.
// Do the remaining pre-update checks and the update while holding a global ontology cache lock.
taskResult <- IriLocker.runWithIriLock(
apiRequestID = deleteOntologyRequest.apiRequestID,
iri = internalOntologyIri.toString,
iri = ONTOLOGY_CACHE_LOCK_IRI,
task = () => makeTaskFuture(
internalOntologyIri = internalOntologyIri
)
Expand Down Expand Up @@ -3147,10 +3148,10 @@ class OntologyResponderV2(responderData: ResponderData) extends Responder(respon
internalPropertyIri = externalPropertyIri.toOntologySchema(InternalSchema)
internalOntologyIri = externalOntologyIri.toOntologySchema(InternalSchema)

// Do the remaining pre-update checks and the update while holding an update lock on the ontology.
// Do the remaining pre-update checks and the update while holding a global ontology cache lock.
taskResult <- IriLocker.runWithIriLock(
apiRequestID = createPropertyRequest.apiRequestID,
iri = internalOntologyIri.toString,
iri = ONTOLOGY_CACHE_LOCK_IRI,
task = () => makeTaskFuture(
internalPropertyIri = internalPropertyIri,
internalOntologyIri = internalOntologyIri
Expand Down Expand Up @@ -3308,10 +3309,10 @@ class OntologyResponderV2(responderData: ResponderData) extends Responder(respon
internalPropertyIri = externalPropertyIri.toOntologySchema(InternalSchema)
internalOntologyIri = externalOntologyIri.toOntologySchema(InternalSchema)

// Do the remaining pre-update checks and the update while holding an update lock on the ontology.
// Do the remaining pre-update checks and the update while holding a global ontology cache lock.
taskResult <- IriLocker.runWithIriLock(
apiRequestID = changePropertyLabelsOrCommentsRequest.apiRequestID,
iri = internalOntologyIri.toString,
iri = ONTOLOGY_CACHE_LOCK_IRI,
task = () => makeTaskFuture(
internalPropertyIri = internalPropertyIri,
internalOntologyIri = internalOntologyIri
Expand Down Expand Up @@ -3430,10 +3431,10 @@ class OntologyResponderV2(responderData: ResponderData) extends Responder(respon
internalClassIri = externalClassIri.toOntologySchema(InternalSchema)
internalOntologyIri = externalOntologyIri.toOntologySchema(InternalSchema)

// Do the remaining pre-update checks and the update while holding an update lock on the ontology.
// Do the remaining pre-update checks and the update while holding a global ontology cache lock.
taskResult <- IriLocker.runWithIriLock(
apiRequestID = changeClassLabelsOrCommentsRequest.apiRequestID,
iri = internalOntologyIri.toString,
iri = ONTOLOGY_CACHE_LOCK_IRI,
task = () => makeTaskFuture(
internalClassIri = internalClassIri,
internalOntologyIri = internalOntologyIri
Expand Down
6 changes: 3 additions & 3 deletions webapi/src/main/scala/org/knora/webapi/util/CacheUtil.scala
View file
Open in desktop
Expand Up @@ -31,7 +31,7 @@ import org.slf4j.LoggerFactory
*/
object CacheUtil {

val log = Logger(LoggerFactory.getLogger("org.knora.webapi.util.cache"))
val log: Logger = Logger(LoggerFactory.getLogger("org.knora.webapi.util.cache"))

/**
* Represents the configuration of a Knora application cache.
Expand All @@ -57,7 +57,7 @@ object CacheUtil {
* `overflowToDisk`, `eternal`, `timeToLiveSeconds`, and `timeToIdleSeconds`,
* representing configuration options for Ehcache caches.
*/
def createCaches(cacheConfigs: Seq[KnoraCacheConfig]) = {
def createCaches(cacheConfigs: Seq[KnoraCacheConfig]): Unit = {
val cacheManager = CacheManager.getInstance()
cacheConfigs.foreach {
cacheConfig =>
Expand All @@ -76,7 +76,7 @@ object CacheUtil {
/**
* Removes all caches.
*/
def removeAllCaches() = {
def removeAllCaches(): Unit = {
val cacheManager = CacheManager.getInstance()
cacheManager.removeAllCaches()
// println("CacheUtil: Removed all application caches")
Expand Down
4 changes: 2 additions & 2 deletions webapi/src/test/scala/org/knora/webapi/responders/IriLockerSpec.scala
View file
Open in desktop
Expand Up @@ -21,7 +21,7 @@ class IriLockerSpec extends WordSpec with Matchers {
"IriLocker" should {
"not allow a request to acquire a lock when another request already has it" in {
def runLongTask(): Future[String] = Future {
Thread.sleep(4500)
Thread.sleep(16000)
SUCCESS
}

Expand Down Expand Up @@ -49,7 +49,7 @@ class IriLockerSpec extends WordSpec with Matchers {
)

val secondTaskFailedWithLockTimeout = try {
Await.result(secondTaskResultFuture, 5.seconds)
Await.result(secondTaskResultFuture, 20.seconds)
false
} catch {
case ale: ApplicationLockException => true
Expand Down