DataONE Tasks: Issueshttps://redmine.dataone.org/https://redmine.dataone.org/favicon.ico2019-06-19T02:03:44ZDataONE Tasks
Redmine Infrastructure - Story #8823 (New): Recent Apache and OpenSSL combinations break connectivity on ...https://redmine.dataone.org/issues/88232019-06-19T02:03:44ZDave Vieglaisdave.vieglais@gmail.com
<p>The latest Ubuntu 18.04 release of Apache is 2.4.29 and OpenSSL is 1.1.1.</p>
<p>This combination creates a significant delay in TLS renegotiation that results from the Apache config option on the CNs:</p>
<pre>SSLVerifyClient none
<Location "/cn">
<If " ! ( %{HTTP_USER_AGENT} =~ /(windows|chrome|mozilla|safari|webkit)/i )">
SSLVerifyClient optional
</If>
</Location>
</pre>
<p>Which is intended to disable client certificate authentication for web browsers, but allow it for others. This approach worked fine on older Apache / OpenSSL but the new combination creates a several second wait when the server discovers the client is not a web browser and tells it to reconnect with the option of including a client certificate.</p>
<p>The latest released version of Apache is 2.4.39 and this is available through a PPA intended for Debian developers. This has been installed so far on dev-2, sandbox, stage, and stage-2 with the process:</p>
<pre>sudo add-apt-repository ppa:ondrej/apache2
sudo apt update
sudo apt dist-upgrade
sudo systemctl restart apache2
</pre>
<p>This installs Apache 2.4.39 and OpenSSL 1.1.1c which appears to resolve the apparent bug in the 2.4.29 / 1.1.1 combination.</p>
<p>One issue with the update is that by default, Apache now offers TLSv1.3, which is great except that it appears to cause problems with at least Python clients failing to connect and getting a 403 error. For example:</p>
<pre>$ python3
>>> import requests
>>> r = requests.get("https://cn-sandbox-ucsb-1.test.dataone.org/cn/v2/monitor/ping")
>>> r.status_code
403
</pre>
<p>That TLSv1.3 is the problem was verified with cn-stage-unm-2 by configuring Apache with:</p>
<pre> SSLProtocol all -TLSv1.3 -SSLv2 -SSLv3
</pre>
<p>to disable TLSv1.3. After this change the Python client was able to connect as expected.</p>
<p>A workaround has not yet been researched.</p>
<p>It is not clear if this issue applies to other clients such as R and Java, so until we learn one way or the other, TLSv1.3 will be disabled on the CNs.</p>
<p>--This issue will likely apply to Member Nodes as well once TLSv1.3 is generally available or if MNs choose to install Apache 2.4.39.-- CORRECTION: this issue only applies when attempting to renegotiate TLS after headers have been transferred, so will not typically apply to a MN.</p>
Infrastructure - Story #8738 (In Progress): HZEventFilter performance decline with increased task...https://redmine.dataone.org/issues/87382018-10-27T16:04:14ZRob Nahfrnahf@epscor.unm.edu
<p>While reindexing, I noticed that creating index tasks was taking about 300ms (when index_task table had about 30k records). Later in the index task generation, that duration increased to about 500 ms on average. (now it's at 600ms).</p>
<p>There are two calls to the database that search for the pid to check its status, and those filters are not against a field that is indexed (pid). Ideally, we should index that field.</p>
<p>At the very least, the 2 queries should be reduced to one query. This could be done without changing the ORM model we're using.</p>
<p>below is the table description in postgres:</p>
<pre>d1-index-queue=# \d index_task
Table "public.index_task"
Column | Type | Collation | Nullable | Default
---------------------+------------------------+-----------+----------+---------
id | bigint | | not null |
datesysmetamodified | bigint | | not null |
deleted | boolean | | not null |
formatid | character varying(255) | | |
nextexecution | bigint | | not null |
objectpath | text | | |
pid | text | | not null |
priority | integer | | not null |
status | character varying(255) | | |
sysmetadata | text | | |
taskmodifieddate | bigint | | not null |
trycount | integer | | not null |
version | integer | | not null |
Indexes:
"index_task_pkey" PRIMARY KEY, btree (id)
d1-index-queue=# \q
</pre> Member Nodes - Story #8535 (In Progress): Neotoma: Story: Discovery & Planninghttps://redmine.dataone.org/issues/85352018-04-09T21:14:39ZAmy Forresteraforres4@utk.edu
<p>Discovery: establish contact and build relationship with a potential new member node. Determine if DataONE and the repository are a good fit for one another and if the repository generally meets the requirements of DataONE member nodes. </p>
<p>Planning: If the repository and DataONE have agreed to proceed with deployment as a member node. Decisions will be made as to how to proceed with development. Node operators will receive training. </p>
<p>This story is complete when a determination is made to either proceed with planning a new deployment, or that joining DataONE is not an option for the repository at this time.</p>
<p><strong>Record initial communication here</strong></p>
Infrastructure - Story #8525 (In Progress): timeout exceptions thrown from Hazelcast disable sync...https://redmine.dataone.org/issues/85252018-03-27T22:36:54ZRob Nahfrnahf@epscor.unm.edu
<p>Very occasionally, synchronization disables itself when RuntimeExceptions bubble up. The most common of these is when the Hazelcast client seemingly disconnects, or can't complete an operation, and a java.util.concurrent.TimeoutException is thrown.</p>
<p>These are usually due to network problems, as evidenced by timeout exceptions appearing in both the Metacat hazelcast-storage.log files as well as d1-processing logs.</p>
<p>Temporary problems like this should be recoverable, and so a retry or bypass for those timeouts should be implemented. It's not clear whether or not a new HazelcastClient should be instantiated, or whether the same client is still usable. (Is the client tightly bound to a session, or does it recover?) If a new client is needed, preliminary searching through the code indicates that refactoring the HazelcastClientFactory.getProcessingClient() method is only used in a few places, and the singleton behavior it uses can be sidestepped by removing the method and replacing it with a getLock() wrapper method (that seems to be the dominant use case for it). See the newer SyncQueueFacade in d1_synchronization for guidance on that. If the client is never exposed, it can be refreshed as needed.</p>
<pre>root@cn-unm-1:/var/metacat/logs# grep FATAL hazelcast-storage.log.1
[FATAL] 2018-03-27 03:15:19,380 (BaseManager$2:run:1402) [64.106.40.6]:5701 [DataONE] Caught error while calling event listener; cause: [CONCURRENT_MAP_CONTAINS_KEY] Operation Timeout (with no response!): 0
</pre><pre>[ERROR] 2018-03-27 03:15:19,781 [ProcessDaemonTask1] (SyncObjectTaskManager:run:84) java.util.concurrent.ExecutionException: java.lang.RuntimeException: java.util.concurrent
.TimeoutException: [CONCURRENT_MAP_REMOVE] Operation Timeout (with no response!): 0
java.util.concurrent.ExecutionException: java.lang.RuntimeException: java.util.concurrent.TimeoutException: [CONCURRENT_MAP_REMOVE] Operation Timeout (with no response!): 0
at java.util.concurrent.FutureTask.report(FutureTask.java:122)
at java.util.concurrent.FutureTask.get(FutureTask.java:192)
at org.dataone.cn.batch.synchronization.SyncObjectTaskManager.run(SyncObjectTaskManager.java:76)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
at java.util.concurrent.FutureTask.run(FutureTask.java:266)
at java.lang.Thread.run(Thread.java:748)
Caused by: java.lang.RuntimeException: java.util.concurrent.TimeoutException: [CONCURRENT_MAP_REMOVE] Operation Timeout (with no response!): 0
at com.hazelcast.impl.ClientServiceException.readData(ClientServiceException.java:63)
at com.hazelcast.nio.Serializer$DataSerializer.read(Serializer.java:104)
at com.hazelcast.nio.Serializer$DataSerializer.read(Serializer.java:79)
at com.hazelcast.nio.AbstractSerializer.toObject(AbstractSerializer.java:121)
at com.hazelcast.nio.AbstractSerializer.toObject(AbstractSerializer.java:156)
at com.hazelcast.client.ClientThreadContext.toObject(ClientThreadContext.java:72)
at com.hazelcast.client.IOUtil.toObject(IOUtil.java:34)
at com.hazelcast.client.ProxyHelper.getValue(ProxyHelper.java:186)
at com.hazelcast.client.ProxyHelper.doOp(ProxyHelper.java:146)
at com.hazelcast.client.ProxyHelper.doOp(ProxyHelper.java:140)
at com.hazelcast.client.QueueClientProxy.innerPoll(QueueClientProxy.java:115)
at com.hazelcast.client.QueueClientProxy.poll(QueueClientProxy.java:111)
at org.dataone.cn.batch.synchronization.type.SyncQueueFacade.poll(SyncQueueFacade.java:231)
at org.dataone.cn.batch.synchronization.tasks.SyncObjectTask.call(SyncObjectTask.java:131)
at org.dataone.cn.batch.synchronization.tasks.SyncObjectTask.call(SyncObjectTask.java:73)
</pre> Member Nodes - Story #8358 (In Progress): Discovery & Planning (CAFF)https://redmine.dataone.org/issues/83582018-02-12T16:25:06ZAmy Forresteraforres4@utk.edu
<p>Discovery is about establishing contact and building a relationship with a potential new member node. In this phase, it is determined if DataONE and the repository are a good fit for one another and if the repository generally meets the requirements of DataONE member nodes. Broad discussions of deployment options may be reviewed as well.<br>
This story is complete when a determination is made to either proceed with planning a new deployment, or that joining DataONE is not an option for the repository at this time.</p>
Infrastructure - Story #8173 (New): add checks for retrograde systemMetadata changeshttps://redmine.dataone.org/issues/81732017-09-01T19:42:33ZRob Nahfrnahf@epscor.unm.edu
<p>with the ability to prioritize and the introduction of parallelized index task processing, the effective queue is not guaranteed to be time-ordered. If there are two valid system metadata changes resulting in two tasks and the second change hits the index first, the earlier task should be rejected, as its changes are out of date.</p>
Python GMN - Story #8032 (In Progress): GMN v2https://redmine.dataone.org/issues/80322017-03-02T18:25:18ZRoger Dahldahl@unm.edu
<p>GMN v2 related tasks that don't require their own story.</p>
Infrastructure - Story #7832 (In Progress): migrate from JibX to JAXB for XML binding / codegenhttps://redmine.dataone.org/issues/78322016-06-24T19:12:07ZRob Nahfrnahf@epscor.unm.edu
<p>JibX is losing support in the community - not finding willing partners for maintainance - while JAXB has become the standard XML binding framework for Java. We plan to migrate / switch to JAXB for object un/marshalling and most likely datatype code generation.</p>
Infrastructure - Story #7358 (In Progress): ContactSubject on NodeList must be valid D1 ldap entryhttps://redmine.dataone.org/issues/73582015-09-16T19:58:13ZRobert Waltz
<p>Before a CN can be started, LDAP must have an approved entry for Contact Subject.</p>
<p>Contact Subject has been defaulted to CN=Robert P Waltz A904,O=Google,C=US,DC=cilogon,DC=org on all of the CN entries in the node list.</p>
<p>Since Robert P Waltz is a developer and not an organizer or director, then the publicized contact on the CNs should be changed to reflect the organizational hierarchy.</p>
<p>The Contact Subject for the CNs should be the PI of the project, or at least, a Co-PI.</p>
<p>Also, The DN of this subject should be derived from the DataONE CA instead of cilogon.</p>
<p>Updating the existing systems should be trivial. The Ldap Entry for each CN node will be modified, and a new LDAP entry for the new Subject will need to be added.</p>
Python GMN - Story #7219 (In Progress): Upgrade Django version used by GMN, add better support fo...https://redmine.dataone.org/issues/72192015-06-16T14:56:59ZDave Vieglaisdave.vieglais@gmail.comOGC-Slender Node - Story #7151 (In Progress): Generate system metadata for contenthttps://redmine.dataone.org/issues/71512015-06-04T20:24:16ZDave Vieglaisdave.vieglais@gmail.com
<p>For the three types of content (resource, science meta, data), develop a mechanism to provide system metadata.</p>
OGC-Slender Node - Story #7149 (Testing): Implement mechanism to retrieve a list of objects avail...https://redmine.dataone.org/issues/71492015-06-04T20:20:47ZDave Vieglaisdave.vieglais@gmail.com
<p>Using Python, implement a tool that is able to retrieve a list of packages, and the objects that make up each package.</p>
OGC-Slender Node - Story #7146 (In Progress): Determine formatId for content retrievable from NOD...https://redmine.dataone.org/issues/71462015-06-04T20:15:08ZDave Vieglaisdave.vieglais@gmail.com
<p>In order to create system metadata for objects held by NODC, it is necessary to infer the appropriate formatId for each object.</p>
Member Nodes - Story #5833 (In Progress): GBIF: Developinghttps://redmine.dataone.org/issues/58332014-07-17T21:05:32ZRoger Dahldahl@unm.edu
<p>Determine which software stack to use, etc.</p>
Infrastructure - Story #4052 (In Progress): OPeNDAP MN Storyhttps://redmine.dataone.org/issues/40522013-10-06T20:07:50ZBruce Wilsonbwilso27@utk.edu
<p>There are a large number of possible DataONE MN's running Data Access Protocol (DAP) compliant servers, particularly servers based on the OPeNDAP Hyrax and UCAR THREDDS servers. The objective of this work is to develop a MN stack that can be used with at least one of these DAP-compliant software stacks as a low-barrier route to becoming a Tier 1 MN.</p>