public void processGetVersions(ResponseBuilder rb) throws IOException {
SolrQueryRequest req = rb.req;
SolrQueryResponse rsp = rb.rsp;
SolrParams params = req.getParams();
if (!params.getBool(COMPONENT_NAME, true)) {
return;
}
int nVersions = params.getInt("getVersions", -1);
if (nVersions == -1) return;
String sync = params.get("sync");
if (sync != null) {
processSync(rb, nVersions, sync);
return;
}
UpdateLog ulog = req.getCore().getUpdateHandler().getUpdateLog();
if (ulog == null) return;
UpdateLog.RecentUpdates recentUpdates = ulog.getRecentUpdates();
try {
rb.rsp.add("versions", recentUpdates.getVersions(nVersions));
} finally {
recentUpdates.close(); // cache this somehow?
}
}
public void processGetUpdates(ResponseBuilder rb) throws IOException {
SolrQueryRequest req = rb.req;
SolrQueryResponse rsp = rb.rsp;
SolrParams params = req.getParams();
if (!params.getBool(COMPONENT_NAME, true)) {
return;
}
String versionsStr = params.get("getUpdates");
if (versionsStr == null) return;
UpdateLog ulog = req.getCore().getUpdateHandler().getUpdateLog();
if (ulog == null) return;
List<String> versions = StrUtils.splitSmart(versionsStr, ",", true);
List<Object> updates = new ArrayList<Object>(versions.size());
long minVersion = Long.MAX_VALUE;
// TODO: get this from cache instead of rebuilding?
UpdateLog.RecentUpdates recentUpdates = ulog.getRecentUpdates();
try {
for (String versionStr : versions) {
long version = Long.parseLong(versionStr);
try {
Object o = recentUpdates.lookup(version);
if (o == null) continue;
if (version > 0) {
minVersion = Math.min(minVersion, version);
}
// TODO: do any kind of validation here?
updates.add(o);
} catch (SolrException e) {
log.warn("Exception reading log for updates", e);
} catch (ClassCastException e) {
log.warn("Exception reading log for updates", e);
}
}
// Must return all delete-by-query commands that occur after the first add requested
// since they may apply.
updates.addAll(recentUpdates.getDeleteByQuery(minVersion));
rb.rsp.add("updates", updates);
} finally {
recentUpdates.close(); // cache this somehow?
}
}
@Override
public void processCommit(CommitUpdateCommand cmd) throws IOException {
if (zkEnabled) {
zkCheck();
}
if (vinfo != null) {
vinfo.lockForUpdate();
}
try {
if (ulog == null
|| ulog.getState() == UpdateLog.State.ACTIVE
|| (cmd.getFlags() & UpdateCommand.REPLAY) != 0) {
super.processCommit(cmd);
} else {
log.info(
"Ignoring commit while not ACTIVE - state: "
+ ulog.getState()
+ " replay:"
+ (cmd.getFlags() & UpdateCommand.REPLAY));
}
} finally {
if (vinfo != null) {
vinfo.unlockForUpdate();
}
}
// TODO: we should consider this? commit everyone in the current collection
if (zkEnabled) {
ModifiableSolrParams params = new ModifiableSolrParams(filterParams(req.getParams()));
if (!req.getParams().getBool(COMMIT_END_POINT, false)) {
params.set(COMMIT_END_POINT, true);
String nodeName =
req.getCore().getCoreDescriptor().getCoreContainer().getZkController().getNodeName();
String shardZkNodeName = nodeName + "_" + req.getCore().getName();
List<Node> nodes =
getCollectionUrls(
req,
req.getCore().getCoreDescriptor().getCloudDescriptor().getCollectionName(),
shardZkNodeName);
if (nodes != null) {
cmdDistrib.distribCommit(cmd, nodes, params);
finish();
}
}
}
}
public static SolrInputDocument getInputDocument(SolrCore core, BytesRef idBytes)
throws IOException {
SolrInputDocument sid = null;
RefCounted<SolrIndexSearcher> searcherHolder = null;
try {
SolrIndexSearcher searcher = null;
UpdateLog ulog = core.getUpdateHandler().getUpdateLog();
if (ulog != null) {
Object o = ulog.lookup(idBytes);
if (o != null) {
// should currently be a List<Oper,Ver,Doc/Id>
List entry = (List) o;
assert entry.size() >= 3;
int oper = (Integer) entry.get(0) & UpdateLog.OPERATION_MASK;
switch (oper) {
case UpdateLog.ADD:
sid = (SolrInputDocument) entry.get(entry.size() - 1);
break;
case UpdateLog.DELETE:
return null;
default:
throw new SolrException(
SolrException.ErrorCode.SERVER_ERROR, "Unknown Operation! " + oper);
}
}
}
if (sid == null) {
// didn't find it in the update log, so it should be in the newest searcher opened
if (searcher == null) {
searcherHolder = core.getRealtimeSearcher();
searcher = searcherHolder.get();
}
// SolrCore.verbose("RealTimeGet using searcher ", searcher);
SchemaField idField = core.getLatestSchema().getUniqueKeyField();
int docid = searcher.getFirstMatch(new Term(idField.getName(), idBytes));
if (docid < 0) return null;
StoredDocument luceneDocument = searcher.doc(docid);
sid = toSolrInputDocument(luceneDocument, core.getLatestSchema());
}
} finally {
if (searcherHolder != null) {
searcherHolder.decref();
}
}
return sid;
}
public DistributedUpdateProcessor(
SolrQueryRequest req, SolrQueryResponse rsp, UpdateRequestProcessor next) {
super(next);
this.rsp = rsp;
this.next = next;
this.idField = req.getSchema().getUniqueKeyField();
// version init
this.updateHandler = req.getCore().getUpdateHandler();
this.ulog = updateHandler.getUpdateLog();
this.vinfo = ulog == null ? null : ulog.getVersionInfo();
versionsStored = this.vinfo != null && this.vinfo.getVersionField() != null;
returnVersions = req.getParams().getBool(UpdateParams.VERSIONS, false);
// TODO: better way to get the response, or pass back info to it?
SolrRequestInfo reqInfo = returnVersions ? SolrRequestInfo.getRequestInfo() : null;
this.req = req;
CoreDescriptor coreDesc = req.getCore().getCoreDescriptor();
this.zkEnabled = coreDesc.getCoreContainer().isZooKeeperAware();
zkController = req.getCore().getCoreDescriptor().getCoreContainer().getZkController();
if (zkEnabled) {
numNodes = zkController.getZkStateReader().getClusterState().getLiveNodes().size();
cmdDistrib =
new SolrCmdDistributor(
numNodes, coreDesc.getCoreContainer().getZkController().getCmdDistribExecutor());
}
// this.rsp = reqInfo != null ? reqInfo.getRsp() : null;
cloudDesc = coreDesc.getCloudDescriptor();
if (cloudDesc != null) {
collection = cloudDesc.getCollectionName();
}
}
@Test
public void testBufferingFlags() throws Exception {
DirectUpdateHandler2.commitOnClose = false;
final Semaphore logReplayFinish = new Semaphore(0);
UpdateLog.testing_logReplayFinishHook =
new Runnable() {
@Override
public void run() {
logReplayFinish.release();
}
};
SolrQueryRequest req = req();
UpdateHandler uhandler = req.getCore().getUpdateHandler();
UpdateLog ulog = uhandler.getUpdateLog();
try {
clearIndex();
assertU(commit());
assertEquals(UpdateLog.State.ACTIVE, ulog.getState());
ulog.bufferUpdates();
// simulate updates from a leader
updateJ(
jsonAdd(sdoc("id", "Q1", "_version_", "101")), params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonAdd(sdoc("id", "Q2", "_version_", "102")), params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonAdd(sdoc("id", "Q3", "_version_", "103")), params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
assertEquals(UpdateLog.State.BUFFERING, ulog.getState());
req.close();
h.close();
createCore();
req = req();
uhandler = req.getCore().getUpdateHandler();
ulog = uhandler.getUpdateLog();
logReplayFinish.acquire(); // wait for replay to finish
assertTrue(
(ulog.getStartingOperation() & UpdateLog.FLAG_GAP)
!= 0); // since we died while buffering, we should see this last
//
// Try again to ensure that the previous log replay didn't wipe out our flags
//
req.close();
h.close();
createCore();
req = req();
uhandler = req.getCore().getUpdateHandler();
ulog = uhandler.getUpdateLog();
assertTrue((ulog.getStartingOperation() & UpdateLog.FLAG_GAP) != 0);
// now do some normal non-buffered adds
updateJ(
jsonAdd(sdoc("id", "Q4", "_version_", "114")), params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonAdd(sdoc("id", "Q5", "_version_", "115")), params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonAdd(sdoc("id", "Q6", "_version_", "116")), params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
assertU(commit());
req.close();
h.close();
createCore();
req = req();
uhandler = req.getCore().getUpdateHandler();
ulog = uhandler.getUpdateLog();
assertTrue((ulog.getStartingOperation() & UpdateLog.FLAG_GAP) == 0);
ulog.bufferUpdates();
// simulate receiving no updates
ulog.applyBufferedUpdates();
updateJ(
jsonAdd(sdoc("id", "Q7", "_version_", "117")),
params(
DISTRIB_UPDATE_PARAM,
FROM_LEADER)); // do another add to make sure flags are back to normal
req.close();
h.close();
createCore();
req = req();
uhandler = req.getCore().getUpdateHandler();
ulog = uhandler.getUpdateLog();
assertTrue((ulog.getStartingOperation() & UpdateLog.FLAG_GAP) == 0); // check flags on Q7
logReplayFinish.acquire();
assertEquals(
UpdateLog.State.ACTIVE, ulog.getState()); // leave each test method in a good state
} finally {
DirectUpdateHandler2.commitOnClose = true;
UpdateLog.testing_logReplayHook = null;
UpdateLog.testing_logReplayFinishHook = null;
req().close();
}
}
@Test
@Ignore("HDFS-3107: no truncate support yet")
public void testDropBuffered() throws Exception {
DirectUpdateHandler2.commitOnClose = false;
final Semaphore logReplay = new Semaphore(0);
final Semaphore logReplayFinish = new Semaphore(0);
UpdateLog.testing_logReplayHook =
new Runnable() {
@Override
public void run() {
try {
assertTrue(logReplay.tryAcquire(timeout, TimeUnit.SECONDS));
} catch (Exception e) {
throw new RuntimeException(e);
}
}
};
UpdateLog.testing_logReplayFinishHook =
new Runnable() {
@Override
public void run() {
logReplayFinish.release();
}
};
SolrQueryRequest req = req();
UpdateHandler uhandler = req.getCore().getUpdateHandler();
UpdateLog ulog = uhandler.getUpdateLog();
try {
clearIndex();
assertU(commit());
assertEquals(UpdateLog.State.ACTIVE, ulog.getState());
ulog.bufferUpdates();
assertEquals(UpdateLog.State.BUFFERING, ulog.getState());
Future<UpdateLog.RecoveryInfo> rinfoFuture = ulog.applyBufferedUpdates();
assertTrue(rinfoFuture == null);
assertEquals(UpdateLog.State.ACTIVE, ulog.getState());
ulog.bufferUpdates();
assertEquals(UpdateLog.State.BUFFERING, ulog.getState());
// simulate updates from a leader
updateJ(
jsonAdd(sdoc("id", "C1", "_version_", "101")), params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonAdd(sdoc("id", "C2", "_version_", "102")), params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonAdd(sdoc("id", "C3", "_version_", "103")), params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
assertTrue(ulog.dropBufferedUpdates());
ulog.bufferUpdates();
updateJ(
jsonAdd(sdoc("id", "C4", "_version_", "104")), params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonAdd(sdoc("id", "C5", "_version_", "105")), params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
logReplay.release(1000);
rinfoFuture = ulog.applyBufferedUpdates();
UpdateLog.RecoveryInfo rinfo = rinfoFuture.get();
assertEquals(2, rinfo.adds);
assertJQ(req("qt", "/get", "getVersions", "2"), "=={'versions':[105,104]}");
// this time add some docs first before buffering starts (so tlog won't be at pos 0)
updateJ(
jsonAdd(sdoc("id", "C100", "_version_", "200")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonAdd(sdoc("id", "C101", "_version_", "201")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
ulog.bufferUpdates();
updateJ(
jsonAdd(sdoc("id", "C103", "_version_", "203")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonAdd(sdoc("id", "C104", "_version_", "204")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
assertTrue(ulog.dropBufferedUpdates());
ulog.bufferUpdates();
updateJ(
jsonAdd(sdoc("id", "C105", "_version_", "205")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonAdd(sdoc("id", "C106", "_version_", "206")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
rinfoFuture = ulog.applyBufferedUpdates();
rinfo = rinfoFuture.get();
assertEquals(2, rinfo.adds);
assertJQ(
req("q", "*:*", "sort", "_version_ asc", "fl", "id,_version_"),
"/response/docs==["
+ "{'id':'C4','_version_':104}"
+ ",{'id':'C5','_version_':105}"
+ ",{'id':'C100','_version_':200}"
+ ",{'id':'C101','_version_':201}"
+ ",{'id':'C105','_version_':205}"
+ ",{'id':'C106','_version_':206}"
+ "]");
assertJQ(req("qt", "/get", "getVersions", "6"), "=={'versions':[206,205,201,200,105,104]}");
ulog.bufferUpdates();
assertEquals(UpdateLog.State.BUFFERING, ulog.getState());
updateJ(
jsonAdd(sdoc("id", "C301", "_version_", "998")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonAdd(sdoc("id", "C302", "_version_", "999")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
assertTrue(ulog.dropBufferedUpdates());
// make sure we can overwrite with a lower version
// TODO: is this functionality needed?
updateJ(
jsonAdd(sdoc("id", "C301", "_version_", "301")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonAdd(sdoc("id", "C302", "_version_", "302")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
assertU(commit());
assertJQ(req("qt", "/get", "getVersions", "2"), "=={'versions':[302,301]}");
assertJQ(
req("q", "*:*", "sort", "_version_ desc", "fl", "id,_version_", "rows", "2"),
"/response/docs==["
+ "{'id':'C302','_version_':302}"
+ ",{'id':'C301','_version_':301}"
+ "]");
updateJ(
jsonAdd(sdoc("id", "C2", "_version_", "302")), params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
assertEquals(
UpdateLog.State.ACTIVE, ulog.getState()); // leave each test method in a good state
} finally {
DirectUpdateHandler2.commitOnClose = true;
UpdateLog.testing_logReplayHook = null;
UpdateLog.testing_logReplayFinishHook = null;
req().close();
}
}
@Test
public void testBuffering() throws Exception {
DirectUpdateHandler2.commitOnClose = false;
final Semaphore logReplay = new Semaphore(0);
final Semaphore logReplayFinish = new Semaphore(0);
UpdateLog.testing_logReplayHook =
new Runnable() {
@Override
public void run() {
try {
assertTrue(logReplay.tryAcquire(timeout, TimeUnit.SECONDS));
} catch (Exception e) {
throw new RuntimeException(e);
}
}
};
UpdateLog.testing_logReplayFinishHook =
new Runnable() {
@Override
public void run() {
logReplayFinish.release();
}
};
SolrQueryRequest req = req();
UpdateHandler uhandler = req.getCore().getUpdateHandler();
UpdateLog ulog = uhandler.getUpdateLog();
try {
clearIndex();
assertU(commit());
assertEquals(UpdateLog.State.ACTIVE, ulog.getState());
ulog.bufferUpdates();
assertEquals(UpdateLog.State.BUFFERING, ulog.getState());
Future<UpdateLog.RecoveryInfo> rinfoFuture = ulog.applyBufferedUpdates();
assertTrue(rinfoFuture == null);
assertEquals(UpdateLog.State.ACTIVE, ulog.getState());
ulog.bufferUpdates();
assertEquals(UpdateLog.State.BUFFERING, ulog.getState());
// simulate updates from a leader
updateJ(
jsonAdd(sdoc("id", "B1", "_version_", "1010")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonAdd(sdoc("id", "B11", "_version_", "1015")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonDelQ("id:B1 id:B11 id:B2 id:B3"),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER, "_version_", "-1017"));
updateJ(
jsonAdd(sdoc("id", "B2", "_version_", "1020")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonAdd(sdoc("id", "B3", "_version_", "1030")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
deleteAndGetVersion("B1", params(DISTRIB_UPDATE_PARAM, FROM_LEADER, "_version_", "-2010"));
assertJQ(
req("qt", "/get", "getVersions", "6"),
"=={'versions':[-2010,1030,1020,-1017,1015,1010]}");
assertU(commit());
assertJQ(
req("qt", "/get", "getVersions", "6"),
"=={'versions':[-2010,1030,1020,-1017,1015,1010]}");
// updates should be buffered, so we should not see any results yet.
assertJQ(req("q", "*:*"), "/response/numFound==0");
// real-time get should also not show anything (this could change in the future,
// but it's currently used for validating version numbers too, so it would
// be bad for updates to be visible if we're just buffering.
assertJQ(req("qt", "/get", "id", "B3"), "=={'doc':null}");
rinfoFuture = ulog.applyBufferedUpdates();
assertTrue(rinfoFuture != null);
assertEquals(UpdateLog.State.APPLYING_BUFFERED, ulog.getState());
logReplay.release(1000);
UpdateLog.RecoveryInfo rinfo = rinfoFuture.get();
assertEquals(UpdateLog.State.ACTIVE, ulog.getState());
assertJQ(
req("qt", "/get", "getVersions", "6"),
"=={'versions':[-2010,1030,1020,-1017,1015,1010]}");
assertJQ(req("q", "*:*"), "/response/numFound==2");
// move back to recovering
ulog.bufferUpdates();
assertEquals(UpdateLog.State.BUFFERING, ulog.getState());
Long ver = getVer(req("qt", "/get", "id", "B3"));
assertEquals(1030L, ver.longValue());
// add a reordered doc that shouldn't overwrite one in the index
updateJ(
jsonAdd(sdoc("id", "B3", "_version_", "3")), params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
// reorder two buffered updates
updateJ(
jsonAdd(sdoc("id", "B4", "_version_", "1040")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
deleteAndGetVersion(
"B4",
params(
DISTRIB_UPDATE_PARAM,
FROM_LEADER,
"_version_",
"-940")); // this update should not take affect
updateJ(
jsonAdd(sdoc("id", "B6", "_version_", "1060")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonAdd(sdoc("id", "B5", "_version_", "1050")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
updateJ(
jsonAdd(sdoc("id", "B8", "_version_", "1080")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
// test that delete by query is at least buffered along with everything else so it will delete
// the
// currently buffered id:8 (even if it doesn't currently support versioning)
updateJ(
"{\"delete\": { \"query\":\"id:B2 OR id:B8\" }}",
params(DISTRIB_UPDATE_PARAM, FROM_LEADER, "_version_", "-3000"));
assertJQ(
req("qt", "/get", "getVersions", "13"),
"=={'versions':[-3000,1080,1050,1060,-940,1040,3,-2010,1030,1020,-1017,1015,1010]}" // the
// "3"
// appears because versions aren't checked while buffering
);
logReplay.drainPermits();
rinfoFuture = ulog.applyBufferedUpdates();
assertTrue(rinfoFuture != null);
assertEquals(UpdateLog.State.APPLYING_BUFFERED, ulog.getState());
// apply a single update
logReplay.release(1);
// now add another update
updateJ(
jsonAdd(sdoc("id", "B7", "_version_", "1070")),
params(DISTRIB_UPDATE_PARAM, FROM_LEADER));
// a reordered update that should be dropped
deleteAndGetVersion("B5", params(DISTRIB_UPDATE_PARAM, FROM_LEADER, "_version_", "-950"));
deleteAndGetVersion("B6", params(DISTRIB_UPDATE_PARAM, FROM_LEADER, "_version_", "-2060"));
logReplay.release(1000);
UpdateLog.RecoveryInfo recInfo = rinfoFuture.get();
assertJQ(
req("q", "*:*", "sort", "id asc", "fl", "id,_version_"),
"/response/docs==["
+ "{'id':'B3','_version_':1030}"
+ ",{'id':'B4','_version_':1040}"
+ ",{'id':'B5','_version_':1050}"
+ ",{'id':'B7','_version_':1070}"
+ "]");
assertEquals(1, recInfo.deleteByQuery);
assertEquals(
UpdateLog.State.ACTIVE, ulog.getState()); // leave each test method in a good state
} finally {
DirectUpdateHandler2.commitOnClose = true;
UpdateLog.testing_logReplayHook = null;
UpdateLog.testing_logReplayFinishHook = null;
req().close();
}
}
private boolean versionDelete(DeleteUpdateCommand cmd) throws IOException {
BytesRef idBytes = cmd.getIndexedId();
if (vinfo == null || idBytes == null) {
super.processDelete(cmd);
return false;
}
// This is only the hash for the bucket, and must be based only on the uniqueKey (i.e. do not
// use a pluggable hash here)
int bucketHash = Hash.murmurhash3_x86_32(idBytes.bytes, idBytes.offset, idBytes.length, 0);
// at this point, there is an update we need to try and apply.
// we may or may not be the leader.
// Find the version
long versionOnUpdate = cmd.getVersion();
if (versionOnUpdate == 0) {
String versionOnUpdateS = req.getParams().get(VERSION_FIELD);
versionOnUpdate = versionOnUpdateS == null ? 0 : Long.parseLong(versionOnUpdateS);
}
long signedVersionOnUpdate = versionOnUpdate;
versionOnUpdate = Math.abs(versionOnUpdate); // normalize to positive version
boolean isReplay = (cmd.getFlags() & UpdateCommand.REPLAY) != 0;
boolean leaderLogic = isLeader && !isReplay;
if (!leaderLogic && versionOnUpdate == 0) {
throw new SolrException(ErrorCode.BAD_REQUEST, "missing _version_ on update from leader");
}
VersionBucket bucket = vinfo.bucket(bucketHash);
vinfo.lockForUpdate();
try {
synchronized (bucket) {
if (versionsStored) {
long bucketVersion = bucket.highest;
if (leaderLogic) {
if (signedVersionOnUpdate != 0) {
Long lastVersion = vinfo.lookupVersion(cmd.getIndexedId());
long foundVersion = lastVersion == null ? -1 : lastVersion;
if ((signedVersionOnUpdate == foundVersion)
|| (signedVersionOnUpdate < 0 && foundVersion < 0)
|| (signedVersionOnUpdate == 1 && foundVersion > 0)) {
// we're ok if versions match, or if both are negative (all missing docs are equal),
// or if cmd
// specified it must exist (versionOnUpdate==1) and it does.
} else {
throw new SolrException(
ErrorCode.CONFLICT,
"version conflict for "
+ cmd.getId()
+ " expected="
+ signedVersionOnUpdate
+ " actual="
+ foundVersion);
}
}
long version = vinfo.getNewClock();
cmd.setVersion(-version);
bucket.updateHighest(version);
} else {
cmd.setVersion(-versionOnUpdate);
if (ulog.getState() != UpdateLog.State.ACTIVE
&& (cmd.getFlags() & UpdateCommand.REPLAY) == 0) {
// we're not in an active state, and this update isn't from a replay, so buffer it.
cmd.setFlags(cmd.getFlags() | UpdateCommand.BUFFERING);
ulog.delete(cmd);
return true;
}
// if we aren't the leader, then we need to check that updates were not re-ordered
if (bucketVersion != 0 && bucketVersion < versionOnUpdate) {
// we're OK... this update has a version higher than anything we've seen
// in this bucket so far, so we know that no reordering has yet occured.
bucket.updateHighest(versionOnUpdate);
} else {
// there have been updates higher than the current update. we need to check
// the specific version for this id.
Long lastVersion = vinfo.lookupVersion(cmd.getIndexedId());
if (lastVersion != null && Math.abs(lastVersion) >= versionOnUpdate) {
// This update is a repeat, or was reordered. We need to drop this update.
return true;
}
}
}
}
doLocalDelete(cmd);
return false;
} // end synchronized (bucket)
} finally {
vinfo.unlockForUpdate();
}
}
public void doDeleteByQuery(DeleteUpdateCommand cmd) throws IOException {
// even in non zk mode, tests simulate updates from a leader
if (!zkEnabled) {
isLeader = getNonZkLeaderAssumption(req);
} else {
zkCheck();
}
// NONE: we are the first to receive this deleteByQuery
// - it must be forwarded to the leader of every shard
// TO: we are a leader receiving a forwarded deleteByQuery... we must:
// - block all updates (use VersionInfo)
// - flush *all* updates going to our replicas
// - forward the DBQ to our replicas and wait for the response
// - log + execute the local DBQ
// FROM: we are a replica receiving a DBQ from our leader
// - log + execute the local DBQ
DistribPhase phase = DistribPhase.parseParam(req.getParams().get(DISTRIB_UPDATE_PARAM));
if (zkEnabled && DistribPhase.NONE == phase) {
boolean leaderForAnyShard = false; // start off by assuming we are not a leader for any shard
Map<String, Slice> slices = zkController.getClusterState().getSlices(collection);
if (slices == null) {
throw new SolrException(
ErrorCode.BAD_REQUEST,
"Cannot find collection:"
+ collection
+ " in "
+ zkController.getClusterState().getCollections());
}
ModifiableSolrParams params = new ModifiableSolrParams(filterParams(req.getParams()));
params.set(DISTRIB_UPDATE_PARAM, DistribPhase.TOLEADER.toString());
List<Node> leaders = new ArrayList<Node>(slices.size());
for (Map.Entry<String, Slice> sliceEntry : slices.entrySet()) {
String sliceName = sliceEntry.getKey();
ZkNodeProps leaderProps;
try {
leaderProps = zkController.getZkStateReader().getLeaderProps(collection, sliceName);
} catch (InterruptedException e) {
throw new SolrException(
ErrorCode.SERVICE_UNAVAILABLE, "Exception finding leader for shard " + sliceName, e);
}
// TODO: What if leaders changed in the meantime?
// should we send out slice-at-a-time and if a node returns "hey, I'm not a leader" (or we
// get an error because it went down) then look up the new leader?
// Am I the leader for this slice?
ZkCoreNodeProps coreLeaderProps = new ZkCoreNodeProps(leaderProps);
String leaderNodeName = coreLeaderProps.getCoreNodeName();
String coreName = req.getCore().getName();
String coreNodeName = zkController.getNodeName() + "_" + coreName;
isLeader = coreNodeName.equals(leaderNodeName);
if (isLeader) {
// don't forward to ourself
leaderForAnyShard = true;
} else {
leaders.add(new StdNode(coreLeaderProps));
}
}
params.remove("commit"); // this will be distributed from the local commit
cmdDistrib.distribDelete(cmd, leaders, params);
if (!leaderForAnyShard) {
return;
}
// change the phase to TOLEADER so we look up and forward to our own replicas (if any)
phase = DistribPhase.TOLEADER;
}
List<Node> replicas = null;
if (zkEnabled && DistribPhase.TOLEADER == phase) {
// This core should be a leader
isLeader = true;
replicas = setupRequest();
} else if (DistribPhase.FROMLEADER == phase) {
isLeader = false;
}
if (vinfo == null) {
super.processDelete(cmd);
return;
}
// at this point, there is an update we need to try and apply.
// we may or may not be the leader.
// Find the version
long versionOnUpdate = cmd.getVersion();
if (versionOnUpdate == 0) {
String versionOnUpdateS = req.getParams().get(VERSION_FIELD);
versionOnUpdate = versionOnUpdateS == null ? 0 : Long.parseLong(versionOnUpdateS);
}
versionOnUpdate = Math.abs(versionOnUpdate); // normalize to positive version
boolean isReplay = (cmd.getFlags() & UpdateCommand.REPLAY) != 0;
boolean leaderLogic = isLeader && !isReplay;
if (!leaderLogic && versionOnUpdate == 0) {
throw new SolrException(ErrorCode.BAD_REQUEST, "missing _version_ on update from leader");
}
vinfo.blockUpdates();
try {
if (versionsStored) {
if (leaderLogic) {
long version = vinfo.getNewClock();
cmd.setVersion(-version);
// TODO update versions in all buckets
doLocalDelete(cmd);
} else {
cmd.setVersion(-versionOnUpdate);
if (ulog.getState() != UpdateLog.State.ACTIVE
&& (cmd.getFlags() & UpdateCommand.REPLAY) == 0) {
// we're not in an active state, and this update isn't from a replay, so buffer it.
cmd.setFlags(cmd.getFlags() | UpdateCommand.BUFFERING);
ulog.deleteByQuery(cmd);
return;
}
doLocalDelete(cmd);
}
}
// since we don't know which documents were deleted, the easiest thing to do is to invalidate
// all real-time caches (i.e. UpdateLog) which involves also getting a new version of the
// IndexReader
// (so cache misses will see up-to-date data)
} finally {
vinfo.unblockUpdates();
}
// forward to all replicas
if (leaderLogic && replicas != null) {
ModifiableSolrParams params = new ModifiableSolrParams(filterParams(req.getParams()));
params.set(VERSION_FIELD, Long.toString(cmd.getVersion()));
params.set(DISTRIB_UPDATE_PARAM, DistribPhase.FROMLEADER.toString());
params.set(
"update.from",
ZkCoreNodeProps.getCoreUrl(zkController.getBaseUrl(), req.getCore().getName()));
cmdDistrib.distribDelete(cmd, replicas, params);
cmdDistrib.finish();
}
if (returnVersions && rsp != null) {
if (deleteByQueryResponse == null) {
deleteByQueryResponse = new NamedList<String>();
rsp.add("deleteByQuery", deleteByQueryResponse);
}
deleteByQueryResponse.add(cmd.getQuery(), cmd.getVersion());
}
}
/**
* @return whether or not to drop this cmd
* @throws IOException If there is a low-level I/O error.
*/
private boolean versionAdd(AddUpdateCommand cmd) throws IOException {
BytesRef idBytes = cmd.getIndexedId();
if (vinfo == null || idBytes == null) {
super.processAdd(cmd);
return false;
}
// This is only the hash for the bucket, and must be based only on the uniqueKey (i.e. do not
// use a pluggable hash here)
int bucketHash = Hash.murmurhash3_x86_32(idBytes.bytes, idBytes.offset, idBytes.length, 0);
// at this point, there is an update we need to try and apply.
// we may or may not be the leader.
// Find any existing version in the document
// TODO: don't reuse update commands any more!
long versionOnUpdate = cmd.getVersion();
if (versionOnUpdate == 0) {
SolrInputField versionField = cmd.getSolrInputDocument().getField(VersionInfo.VERSION_FIELD);
if (versionField != null) {
Object o = versionField.getValue();
versionOnUpdate =
o instanceof Number ? ((Number) o).longValue() : Long.parseLong(o.toString());
} else {
// Find the version
String versionOnUpdateS = req.getParams().get(VERSION_FIELD);
versionOnUpdate = versionOnUpdateS == null ? 0 : Long.parseLong(versionOnUpdateS);
}
}
boolean isReplay = (cmd.getFlags() & UpdateCommand.REPLAY) != 0;
boolean leaderLogic = isLeader && !isReplay;
VersionBucket bucket = vinfo.bucket(bucketHash);
vinfo.lockForUpdate();
try {
synchronized (bucket) {
// we obtain the version when synchronized and then do the add so we can ensure that
// if version1 < version2 then version1 is actually added before version2.
// even if we don't store the version field, synchronizing on the bucket
// will enable us to know what version happened first, and thus enable
// realtime-get to work reliably.
// TODO: if versions aren't stored, do we need to set on the cmd anyway for some reason?
// there may be other reasons in the future for a version on the commands
boolean checkDeleteByQueries = false;
if (versionsStored) {
long bucketVersion = bucket.highest;
if (leaderLogic) {
boolean updated = getUpdatedDocument(cmd, versionOnUpdate);
if (versionOnUpdate != 0) {
Long lastVersion = vinfo.lookupVersion(cmd.getIndexedId());
long foundVersion = lastVersion == null ? -1 : lastVersion;
if (versionOnUpdate == foundVersion
|| (versionOnUpdate < 0 && foundVersion < 0)
|| (versionOnUpdate == 1 && foundVersion > 0)) {
// we're ok if versions match, or if both are negative (all missing docs are equal),
// or if cmd
// specified it must exist (versionOnUpdate==1) and it does.
} else {
throw new SolrException(
ErrorCode.CONFLICT,
"version conflict for "
+ cmd.getPrintableId()
+ " expected="
+ versionOnUpdate
+ " actual="
+ foundVersion);
}
}
long version = vinfo.getNewClock();
cmd.setVersion(version);
cmd.getSolrInputDocument().setField(VersionInfo.VERSION_FIELD, version);
bucket.updateHighest(version);
} else {
// The leader forwarded us this update.
cmd.setVersion(versionOnUpdate);
if (ulog.getState() != UpdateLog.State.ACTIVE
&& (cmd.getFlags() & UpdateCommand.REPLAY) == 0) {
// we're not in an active state, and this update isn't from a replay, so buffer it.
cmd.setFlags(cmd.getFlags() | UpdateCommand.BUFFERING);
ulog.add(cmd);
return true;
}
// if we aren't the leader, then we need to check that updates were not re-ordered
if (bucketVersion != 0 && bucketVersion < versionOnUpdate) {
// we're OK... this update has a version higher than anything we've seen
// in this bucket so far, so we know that no reordering has yet occurred.
bucket.updateHighest(versionOnUpdate);
} else {
// there have been updates higher than the current update. we need to check
// the specific version for this id.
Long lastVersion = vinfo.lookupVersion(cmd.getIndexedId());
if (lastVersion != null && Math.abs(lastVersion) >= versionOnUpdate) {
// This update is a repeat, or was reordered. We need to drop this update.
return true;
}
// also need to re-apply newer deleteByQuery commands
checkDeleteByQueries = true;
}
}
}
boolean willDistrib = isLeader && nodes != null && nodes.size() > 0;
SolrInputDocument clonedDoc = null;
if (willDistrib) {
clonedDoc = cmd.solrDoc.deepCopy();
}
// TODO: possibly set checkDeleteByQueries as a flag on the command?
doLocalAdd(cmd);
if (willDistrib) {
cmd.solrDoc = clonedDoc;
}
} // end synchronized (bucket)
} finally {
vinfo.unlockForUpdate();
}
return false;
}
@Override
public void process(ResponseBuilder rb) throws IOException {
SolrQueryRequest req = rb.req;
SolrQueryResponse rsp = rb.rsp;
SolrParams params = req.getParams();
if (!params.getBool(COMPONENT_NAME, true)) {
return;
}
String val = params.get("getVersions");
if (val != null) {
processGetVersions(rb);
return;
}
val = params.get("getUpdates");
if (val != null) {
processGetUpdates(rb);
return;
}
String id[] = params.getParams("id");
String ids[] = params.getParams("ids");
if (id == null && ids == null) {
return;
}
String[] allIds = id == null ? new String[0] : id;
if (ids != null) {
List<String> lst = new ArrayList<String>();
for (String s : allIds) {
lst.add(s);
}
for (String idList : ids) {
lst.addAll(StrUtils.splitSmart(idList, ",", true));
}
allIds = lst.toArray(new String[lst.size()]);
}
SolrCore core = req.getCore();
SchemaField idField = core.getLatestSchema().getUniqueKeyField();
FieldType fieldType = idField.getType();
SolrDocumentList docList = new SolrDocumentList();
UpdateLog ulog = core.getUpdateHandler().getUpdateLog();
RefCounted<SolrIndexSearcher> searcherHolder = null;
DocTransformer transformer = rsp.getReturnFields().getTransformer();
if (transformer != null) {
TransformContext context = new TransformContext();
context.req = req;
transformer.setContext(context);
}
try {
SolrIndexSearcher searcher = null;
BytesRef idBytes = new BytesRef();
for (String idStr : allIds) {
fieldType.readableToIndexed(idStr, idBytes);
if (ulog != null) {
Object o = ulog.lookup(idBytes);
if (o != null) {
// should currently be a List<Oper,Ver,Doc/Id>
List entry = (List) o;
assert entry.size() >= 3;
int oper = (Integer) entry.get(0) & UpdateLog.OPERATION_MASK;
switch (oper) {
case UpdateLog.ADD:
SolrDocument doc =
toSolrDoc(
(SolrInputDocument) entry.get(entry.size() - 1), core.getLatestSchema());
if (transformer != null) {
transformer.transform(doc, -1); // unknown docID
}
docList.add(doc);
break;
case UpdateLog.DELETE:
break;
default:
throw new SolrException(
SolrException.ErrorCode.SERVER_ERROR, "Unknown Operation! " + oper);
}
continue;
}
}
// didn't find it in the update log, so it should be in the newest searcher opened
if (searcher == null) {
searcherHolder = core.getRealtimeSearcher();
searcher = searcherHolder.get();
}
// SolrCore.verbose("RealTimeGet using searcher ", searcher);
int docid = searcher.getFirstMatch(new Term(idField.getName(), idBytes));
if (docid < 0) continue;
StoredDocument luceneDocument = searcher.doc(docid);
SolrDocument doc = toSolrDoc(luceneDocument, core.getLatestSchema());
if (transformer != null) {
transformer.transform(doc, docid);
}
docList.add(doc);
}
} finally {
if (searcherHolder != null) {
searcherHolder.decref();
}
}
// if the client specified a single id=foo, then use "doc":{
// otherwise use a standard doclist
if (ids == null && allIds.length <= 1) {
// if the doc was not found, then use a value of null.
rsp.add("doc", docList.size() > 0 ? docList.get(0) : null);
} else {
docList.setNumFound(docList.size());
rsp.add("response", docList);
}
}
