Guest

Cisco Videoscape Distribution Suite for Internet Streaming

Release Notes for Cisco TV CDS 2.3.3

  • Viewing Options

  • PDF (275.5 KB)
  • Feedback
Release Notes for Cisco TV CDS 2.3.3

Table Of Contents

Release Notes for Cisco TV CDS 2.3.3

Contents

New Features

ISA Environments

RTSP Environments

Multiple Vault Groups

Enhancements

Supported Environments

System Requirements

Important Notes

Open Caveats

Cache2app

Caching Node

Configuration

CServer

Database

FSI

ISA

Monitoring

Network

Other

Streamer

Vault

Resolved Caveats

Cach2app

CServer

Configuration

Database

File Systems

FSI

Installation

ISA

Memory

Monitoring

Network

Other

RTSP

SNMP

statsd

Streamer

Utilities

Vault

Video Quality

Upgrading to Release 2.3.3

Updating the Device Drivers for new SSDs

Rolling Back the SSD Driver Update

Related Documentation

Obtaining Documentation and Submitting a Service Request


Release Notes for Cisco TV CDS 2.3.3


These release notes cover Cisco TV CDS Release 2.3.3.

Revised: October 2011 OL-24132-02

Contents

The following information is in the release notes:

New Features

Enhancements

Supported Environments

System Requirements

Important Notes

Open Caveats

Resolved Caveats

Upgrading to Release 2.3.3

Related Documentation

Obtaining Documentation and Submitting a Service Request

New Features

The following features are supported in Release 2.3:

ISA Environments

RTSP Environments

ISA Environments

ISA environments with a centrally managed VVI supports the following new features:

Real-Time Ingest (RTI) from local Vault Groups.

Steering local live stream ingests to local Vault Groups.

Streamers able to locate content in central (national) Vault Groups for VOD content or a local Vault Group for real-time streaming.

Supports Play Now feature on STBs, which allows a STB to connect to the live point on a live stream and rewind to the beginning.

RTSP Environments

RTSP environments support the following new features:

Playlist support for Quative (off-vpath) with up to four items per playlist.

New Cisco RTSP model (RTSP Deployment Type) that includes interaction with an Authentication Manager and supports both IPTV and cable stream destination types.

EventIS I-30 interface.

Interaction with Zapper set-top box (STB).

Redirect Server—Configuration of the port used for redirecting RTSP requests to another Streamer and configuration of the debug level setting.

Interaction with Cisco Content Delivery System (CDS) Video Navigator and Poster Art Server (PAS) Release 2.3.

Multiple Vault Groups

Up to four Vault Groups are supported in Release 2.3. If Vault Group Redundancy is enabled, content is mirrored among as many as four Vault Groups (one Vault Group ingests the content and up to three Vault Groups mirror the content). Each Vault Group has a separate mirroring policy defined by the Vault Redundancy and the number of copies stored for each Vault in the Vault Group.

In Release 2.3, a Vault Group is configured as the master Vault Group, with one of the Vaults in that group acting as the master Vault. If the master Vault fails, another Vault in that group takes over as the master.


Note Continue using the "nofailover 1" setting in the .arroyorc file on the Vaults that should not takeover as master, if applicable to your configuration.



Note In ISA environments with centrally managed VVIs and Ingest Steering, only central Vault Groups participate in Vault Group Redundancy. Local Vault Groups do not participate in central Vault Group redundancy.


Enhancements

The following enhancements have been added in Release 2.3:

Database Patrol is a new mechanism to keep the database synchronized.

Multiple power supply fan, temperature and failure monitoring.

Supported Environments

Release 2.3.3 of the Cisco TV CDS supports the following environments and associated backoffice integrations:

ISA environment

Tandberg OpenStream backoffice

Onewave backoffice

RTSP environment

SeaChange Axiom backoffice (NGOD architecture)

EventIS

Minerva

Quative

Myrio

Coship

Eyeka

System Requirements

The Cisco TV CDS Release 2.3.3 runs on the CDE110, CDE220, and CDE420. See the Cisco Content Delivery Engine110 Hardware Installation Guide, and the Cisco Content Delivery Engine 205/220/420 Hardware Installation Guide.

The Cisco TV CDS Release 2.3.3 also runs on the CDE100, CDE200, CDE300, and CDE400 hardware models that use the Lindenhurst chipset. See the Cisco Content Delivery Engine CDE100/200/300/400 Hardware Installation Guide for set up and installation procedures.

Release 2.3.3 does not support the CDEs with the ServerWorks chipset. All CDEs with the ServerWorks chipset need to be replaced with the CDEs with the Lindenhurst chipset or the Next Generation Appliances (CDE110, CDE220, andCDE420) before upgrading to Release 2.3.3.

See the "Related Documentation" section for links to the documentation online.

Important Notes

The section covers important notes that apply to this release.

New SSD Driver

There is a new driver for the new solid-state drives (SSDs) that also addresses an issue with the old SSDs.

Old SSD Issue

When an abrupt power loss occurs on an older SSD while the SSD is under load, which can also happen when the SSD is being written to and a power-off shutdown occurs, the result is that the SSD may become corrupted and unusable. If an SSD becomes corrupted in this way it must be replaced.

This driver update provides a fix to prevent the corruption of the SSD when the system is being shutdown. When the system is being shutdown the driver sends a command to the SSD to stop all activity and therefore is protected from corruption.


Note We recommend that this driver update be installed on all systems with SSDs. The driver update is compatible with both the older PV SSDs and the newer PVR SSDs.


New SSDs

New solid-state drives (SSDs) used in the following CDE models require a device driver update for the TV CDS software to recognize the drives:

CDE220-2S1

CDE220-2S3

CDE250-2S5

CDE250-2S6

A new Generation 2 SSD drive could occur in one of the above newly manufactured CDEs or as a returned merchandise authorization (RMA) SSD drive replacement.

The new SSDs replace the end-of-life (EOL) external 160 GB SSDs. A new SSD drive is a Generation 2 front-mounted SSD. The new SSD model is identified by the title "Intel SSD 320 Series" on the label and has the model number: SSDSA2BW160G3. For more information, see Field Notice 63438 at: http://www.cisco.com/en/US/products/ps7126/prod_field_notices_list.html.

The new driver prevents the corruption of the existing SSDs in the field when a system is shutdown. In the case of receiving any new model of SSDs, the driver must be installed.


Note The new driver is 100 percent backward-compatible with older, pre-existing SSD drives. We recommend that a proactive upgrade of the SSD driver is a good best-practice.


For more information, see the "Updating the Device Drivers for new SSDs" section.

Vault Redundancy Mode

When upgrading from 2.1.x and 2.2.x to 2.3.3 with Vault Redundancy Mode turned on, the master Vault Group must be defined; otherwise, no Vault is acting as a master Vault.

Open Caveats

This release contains the following open caveats:

Cache2app

CSCth22711

Symptom:

Server was hard hung.

Conditions:

This can occur with old hardware with a bad controller.

Workaround:

Reboot. Replace the hardware.

CSCth26207

Symptom:

The following message is seen in the /arroyo.log/controlblocktiming.log.<date> file.

Control block CM::RemoteFillWatcher work consumed xx.x percent of the polling process 
CPU for xx seconds
 
   

Conditions:

The disk and mirroring evaluators are disabled. A drive has gone bad and the Vault is trying to repair it. The Vault is queuing thousands of mirroring requests from other Vaults. These requests are not processed until the evaluator is enabled again.

Workaround:

Restart the Vault.

CSCti00738

Symptom:

CalypsoContentStoreServer process on the Vault restarts.

Conditions:

When an ingest provisioning request comes in and the backoffice is unreachable at the time the backoffice is queried for the pre-encryption details. This query is made within a second or two from the initial ingest request.

Workaround:

None.

CSCtk64351

The ISA StreamService keeps getting restarted. The streams were getting stuck, so destroy stream API internally to the ISA/cache2app blocking a thread and eventually after losing 16 threads, the monitor script restarts the ISA StreamService.

Conditions:

There is an issue with the Primary or backup Control streamer that causes streams to get stuck.

Workaround:

Offload the Primary Control streamer and reboot it.

Caching Node

CSCth13373

Symptom:

Caching Node generates a kernel core file and reboots.

Conditions:

Core file caused by a rare multi-CPU race condition in TCP network stack, which results in a divide-by-zero fault.

Workaround:

None.

CSCtk58386

Symptom:

When a disk drive is going bad, duplicate 1x goid copy will be generated. In the meanwhile, a small portion of streams lost fills from Cache Nodes and became open.

Conditions:

The issue is a bad disk drive caused duplication of goid copies and streaming failures. The bad disk drive should be auto detected and permanently removed.

Workaround:

Soft remove the bad disk drive and replace it with good disk drive.

Configuration

CSCti70956

Symptom:

Cannot enter host names containing "." through the CDSM GUI. Error message is displayed preventing entry.

Conditions:

The form field validation does not allow for "." in Host Name field.

Workaround:

Manually edit the host file.

CSCti78219

Symptom:

Not all packages are found in the dropdown on the Monitor > System Level > Package Monitor page.

Conditions:

When the number of packages in the system reaches a level greater than ~70K.

Workaround:

If the number of packages in the system grows much beyond 70K, then the dropdown may not be able to display them all. The solution is to delete packages that are no longer needed. The max number of packages that will display in the dropdown is variable, and is dependent on the system that is accessing the web page.

CSCtk98203

Symptom:

Accessing the groups in Configure->Array Level Stream Groups, Configure-> Array Level -> Vault Groups, Configure -> Array Level -> Cache groups results the GUI page not responding.

Conditions:

This situation may occur if a server that was the part of the group was removed from the CDSM using /home/stats/svrinit_15 -d, without removing it from the group.

Workaround:

1. Add the server that was causing this issue back to the CDSM using /home/stats/svrinit_15 , with the the same groupid and server id as it was having earlier when it was managed by the CDSM.

2. Remove it from the Stream/Vault/Cache Group.

3. Execute /home/stats/svrinit_15 -d to remove the server.

CSCtk98236

Symptom:

On the CDSM GUI Configure > Array Level > Barker Setup page, the number of contents available for selection in the provided dropdown is found to only be 100 contents.

Conditions:

This issue is encountered in both ASI and GigE streaming modes when the system has more than 100 completed ingests.

Workaround:

The API called to return a list of contents available for use in a Barker stream returns the first 100 contents in a lexicographic order. Therefore, it is recommended that customers wishing to stream Barkers name their Barker contents using the prefix "000". For example, the first Barker content would be named "000_barkername1.mpg", the second "000_barkername2.mpg", etc. This will ensure that the content the customer wants available for Barker streams is always available within the drop-down lists on the Configure > Array Level > Barker Setup page.

CServer

CSCth80826

Symptom:

No support for Live Ingest and Streaming for CDN.

Conditions:

Live Streaming of C2 Index formatted files.

Workaround:

This is a new feature scheduled for 2.5.

CSCti02825

Symptom:

Took too long to initialize CServer memory and the system crashed.

Conditions:

Potentially bad memory in the server or bad controller.

Workaround:

Reboot the server or check and replace hardware.

CSCti12423

Symptom:

An input/output request is not returned from the host bus adapter (HBA) driver during a reset. This can cause the evaluators to hang which may have cascading effects to other nodes in the system.

Conditions:

This potentially only occurs with devices that are struggling with input/output requests because of the device or media errors.

Workaround:

None.

CSCti34224

Symptom:

No support for Live Ingest and Streaming for CDN.

Conditions:

Live Streaming of C2 Index formatted files.

Workaround:

This is a new feature scheduled for 2.5.

CSCtj18612

Symptom:

Large number of damaged and/or missing goids.

Conditions:

Create thin pipe, force system out of sync.

Workaround:

If sites are overwhelmingly out of sync the system will take a large amount of time for the evaluators to replicate and clean up data through the thin pipe. Over time, system will clean up.

CSCtk56148

Symptom:

No support for Live Ingest and Streaming for CDN.

Conditions:

Live Streaming of C2 Index formatted files .

Workaround:

This is a new feature scheduled for 2.5.

CSCtk57387

Symptom:

If a streamer server in group A was able to communicate with a server in group B, all streamers in group A show the service address record of the other streamer.

Conditions:

Streamers in group A show the service address record of the other streamer

Workaround:

The code could be changed so that when servers send the service address state to each other, they would not send information for any service addresses for which the sending server has never communicated with either the primary or backup server for that service. This change would automatically stop propagating the service address information to new servers when there are no longer any servers running that have actually communicated with the primary or backup server for that service address.

CSCtk66303

Symptoms:

The cache server has a general protection fault in CM::LocalReadRate::checkAndFillHole+0x66c as can be see in the back trace.

Conditions:

The crash was caused because an acknowledge packet sent from a streamer to the cache server has a bad length for the amount of data acknowledged. This causes the cache server to unlock the wrong amount of data associated with the fill and this causes the later protection fault.

Workaround:

Older streamers are known to cause this problem. Updating streamers to the latest release may resolve the issue. Still looking for an additional cause.

Database

CSCtg93593

Symptom:

Replay.db grows to excessively large size, eventually running the filesystem out of diskspace.

Condition:

An extended outage event such as network partitioning betweenStreamers and Vaults or CDSM can cause a large number of error events (and SNMP traps), which are sent to the CDSM and they become queued on the source device, which causes this issue.

Workaround:

Restore connectivity between sites and servers to avoid buildup of error events. Database stopped on the Brussels streamer due to partition issue.

1. Delete the reply.db (17G) and restart the database.

2. Bounce statsd on both vlt 7 & brx1str.

3. Re-push configs from CDSM to both servers, and verify the files have updated on both servers.

4. Then copy over a new DB to BRX1str.

CSCtk62560

Symptom:

Setup file contains wrong RTSP ports in the setupfile on the Streamer.

Conditions:

RTSP Deployment Type. Centralized VVIM/CDSM or Streamer CDSM side.

Workaround:

Check for the following in the statsd.log.DATE:

"ERROR: no deployment type found, server is NOT configured for isa via bms vendor!" in statsd log of the streamer.

RESOLUTION:

1. Log in as arroyo, change the deployment type from NGOD to something else, submit, and then change it back to NGOD, submit.

2. Log into the various streamers and restart the statsd.

CONFIRM: statsd.log.DATE says "RTSP Deployment type is [2]" on statsd log. (if the customer environment is a NGOD RTSP Env).

CSCtk84103

Symptom:

Database shutdown process took prolonged time for 73 minutes on the master Vault. During shutdown process, database is unavailable on the master Vault.

Conditions:

Database has a big number of assets.

Workaround:

None

CSCtl03347

Symptom:

When viewing the avsdb.log file, messages of the following sort may appear which may seem to indicate an error has occurred:

xx-xx-2010 01:32:50PM: CONTENT_OBJECT_GETALL_RATELIMIT for sockfd=195 starting
xx-xx-2010 01:32:50PM: CONTENT_OBJECT_GETALL_RATELIMIT for sockfd=195 exiting due to 
counter = 1 >= limit = 1
xx-xx-2010 01:32:50PM: CONTENT_OBJECT_GETALL_RATELIMIT for sockfd=195 send 9999 failed
xx-xx-2010 01:32:50PM: CONTENT_OBJECT_GETALL_RATELIMIT for sockfd=195 returning with 
error
 
   

A log of "send 9999 failed" appears after a mention of exiting due to a counter reaching its limit of 1, and then a "returning with error" message for the same sockfd appears.

Conditions:

These messages appear during content object resynchronization, which will also take place during node addition and patrol functionality. The messages would appear on the master Vault/SSV from which another server is performing the content object resynchronization.

Workaround:

This is a cosmetic issue as there is actually no error. The client server makes a test request to the master Vault/SSV in order to check for rate limit support, and then closes the connection before continuing with its actual getall request. This is normal procedure, and these messages simply should not appear in the logs. Please disregard these messages if and when they appear, as they are not indicative of an actual error.

FSI

CSCtk69310

Symptom:

The NGOD A3 live recording TransferStatus asynchronous notification message doesn't contain correct bit rate value. e.g. bitrate = 0.

Conditions:

For NGOD A3 live recording, all the status asynchronous notification has this problem.

Workaround:

After recording completes, the client can actively issue one GetTransferStatus request to get the final bit rate.

ISA

CSCth65408

Symptom:

The stream setup fails continuously if there is only one stream control and it gets stuck. The existing sessions also losses the trick-mode functionality.

Conditions:

The stream control network (QPSK) might have some issue so sending the LSCP response gets blocked.

Workaround:

Restarted the LSCP service as user ISA.

./stop_calypso_service --lscp

Monitoring

CSCtj33259

Symptom:

If none of the vaults are available to perform the ingest then provision request is failed but content is not marked as out of service. This causes reporting issues to the VVIM, the ingest is failed but not reported as failed.

Conditions:

There are some configuration or connectivity issues between vaults causing none of the vaults available to perform ingests.

Workaround:

Problem that is causing none of the vaults available to do ingests should be rectified.

Network

CSCtj57299

Symptom:

When remote mirroring (within an array) or array mirroring (to another array/site) is enabled, the ingest code will attempt to concurrently make a copy of the content to the remote vault(s). However, if the network is unable to keep up with the level of ingest traffic especially when a thin-pipe is present), then the ingest mirroring logic will send different "chunks" of the ingested files to different remote servers. The side effect is that this causes the system to temporarily generate more than the desired number of copies of content. This is clearly inefficient. The fix is to make this ingest mirroring code more intelligent in how it mirrors content. However, this will involve changing the protocol and was deemed too risky for the 2.3.3 release. Eventually, the mirroring logic will "clean-up" these extra copies.

Several issues were discovered due to the fact that this code was generating an excessive number of copies. These defects have been resolved leaving this one as mostly an annoying inefficiency.

Conditions:

Mirroring is not able to keep up with the ingest or the "best" vault changes while ingesting (this is a normal condition).

Workaround:

None.

CSCtj84853

Symptom:

The Stream setups are failing intermittently because the CServer is failing "SetStreamInfo" for a particular control, which is not reachable any more because of connectivity issues on the caching network.

Conditions:

The caching network between Primary setup streamers and Primary Control streamer is partitioned.

Workaround:

There are multiple workarounds.

CSCtk66793

Symptom:

This problem occurs only if stream control interface is different than management interface. If the service is primary or back up control and this interface goes down then cserver does not perform the failover for the control role it has.

Conditions:

Switch or interface port could go down or pulled manually or cable is defective.

Workaround:

Enabled the interface, or offload the server and reboot it.

Other

CSCsq57645

Symptom:

AIM unable to make directories in the /arroyo/asmrpt/adi/packages directory on the CDSM.

Conditions:

Number of directories exceeds 32K which is an ext3 limit.

Workaround:

Remove directories from the /arroyo/asmrpt/adi/packages directory on CDSM. Fix will be included in next major release.

CSCtj22751

Symptom:

Assets won't ingest. If you look at the log you'll see CDS had trouble finding the PMT.

Conditions:

The PMT must span packets, meaning it must occupy more bytes than will fit in a single 188-byte MPEG-2 transport packet. Currently CDS TV does not support such assets. Generally this happens internationally (unlikely in the US) where the extra closed caption streams or extra audio languages can cause this to happen.

Workaround:

Reduce the number of languages or subtitles/captions in the stream when it is encoded.

CSCtj33903

Symptom:

Currently, the steamer waits 500ms for C2 to locate a response which could cause some Content Not Found Error in production environment due to the delay.

Conditions:

C2 locate response takes more than 500ms .

Workaround:

None.

CSCtj47588

Symptom

When a Power Supply (PS) fails, a PS warning alarm is generated, but not if (a redundant) PS is removed.

Conditions:

The PS warning alarm is generated (in the SNMP trap) when the redundant PS looses power from a power source or when the power cord is pulled out. But when the redundant PS is removed totally from the system, it momentarily generates the alarm but disappears soon after and is not seen anymore.

Workaround:

Upgrade the Release Builds to 2.33, 2.42 or later versions.

Streamer

CSCth05766

Symptom:

Evaluator disabled message during Data Recovery.

Conditions:

When there is a very thin pipe defined (for example, less than 100 Mbps) and a very large title count (for example, greater than250,000 titles), the data recovery evaluator can stop working between multiple sites.

Workaround:

Define larger communication pipes.

CSCti79844

Symptom:

This defect typically should not have any adverse affect on a production system. If there are more than eight setup and control service address managers configured in /arroyo/test/streamer/setupfile, including the duplicate, there is the possibility that the final service address entries listed after the eight will be ignored if the are duplicate entries.

Conditions:

This defect is characterized by having duplicate service address entries in the /arroyo/test/setupfile. For example:

service address 10.94.153.130 setup 3300 control 9000
service address 10.94.153.130 setup 3300 control 9000
 
   

Workaround:

Manually remove duplicate service address entries in /arroyo/test/streamer/setupfile, reboot and ensure that the duplicate entries no longer exists after reboot.

CSCti96225

Symptom:

Missing transition event in ODRMSessionHistory.

CSCtj66430

Symptom:

CGW continues to serve cache-fill streams even when it loses connection to all the vaults, which cause streaming errors.

Conditions:

2.2.1 CGW loses connection to all the vaults.

Workaround:

None.

Vault

CSCtf11524

Symptom:

Vault will crash when number of content is above 300,000 titles.

Conditions:

When ingesting more than 300,000 title the system crashes.

Workaround:

Don't ingest more than 300,000 titles.

CSCtf11533

Symptom:

Vault will crash when number of content is above 300,000 titles.

Conditions:

When ingesting more than 300,000 title the system crashes.

Workaround:

Don't ingest more than 300,000 titles.

CSCtk34612

Symptom:

The logic for cache-fill needs to consider unequal thin pipe setting and bandwidth analysis.

Conditions:

This behavior was noticed while trying to debug an issue where a customer was seeing an imbalance in traffic between multiple sites. After further analysis it was determined that this issue was not the actual cause of the extreme imbalance. Further review determined that this condition has only minor affects given all the other dynamics in a multiple site system.

Workaround:

None.

Resolved Caveats

The following caveats have been resolved since Cisco TV CDS Release 2.3.3. Not all resolved issues are mentioned here. The following list highlights associated with customer deployment scenarios.

Cach2app

CSCtg18769

Symptom:

When adding or deleting a stream destination from the CDSM GUI, if the master Streamer goes into kernel debugger (KDB) mode, the CalypsoClient process is stuck at the CDSM and the script to add or delete a new stream destination for that Service Group is not able to run.

CServer

CSCth31829

Symptom:

Mirroring within an array is slow due to excessive memory being consumed by communications to another array.

Conditions:

Mirroring is completely blocked because the thin pipe settings are not allowing data to flow because the per interval value is less than a full packet.

CSCth87732

Symptom:

The "PAID not found: hbo.com::HBOD1000000000575947::index status: 4" type error messages need to print out the vault to which the CGW communicated. This is to help the debugging process.

CSCti94507

Symptom:

Vault is constantly coring.

CSCtj02140

Symptom:

When rebooting C2 master, all the CGW in the cache group started seeing, unseeing, each other per the CServer.

CSCtj16184

Symptom:

Hardware Monitors reporting improper values on some servers inconsistently.

CSCtj77501

Symptom:

CGW core during cache filling and experience network connectivity issue.

Configuration

CSCtg88234

Symptom:

Submitting System-> Headend Setup page removes the leading zeros in the Service Group Name.

CSCth21572

Symptom:

A device (vault, cache or streamer) in a VVIM may accidentally delete the /arroyo/test/RemoteServers file while doing a system reconfiguration. This reconfiguration may be triggered in multiple ways, such as adding new Vaults. Without an appropriate RemoteServers file, the affected device cannot communicate with other devices in the network.

CSCth69300

Symptom:

New DSCP markings get configured in the /arroyo/test/cache/setupfile and /arroyo/test/vault/setupfile. The values are selected from the Configure > Server Level > Server Setup page, under Assured Forwarding Class. There is no way to keep these settings out of the setupfile.

CSCti03224

Symptom:

All servers on the platform get unregistered from their respective groups (stream/vault/cache/ssv).

SCti17154

Symptom:

The RemoteServers contains all streamers in a deployment, not simply the streamers that are part of the group.

CSCtj77728

Symptom:

Some changes that can be made to the subnet table require a reboot. The GUI needs to alert the user to this fact.

CSCtk30422

Symptom:

Peer Entries are removed from the remote sever file.

Database

CSCtc36567

Symptom:

Deleted content can sometimes be successfully deleted in the CServer and database; however, in the content reports they are shown with a delete timestamp of N/A. This is due to the VVIM being down for a longer period of time and upon restoration of the VVIM the content objects are resynchronized. It is during this resynchronization process that the delete timestamp is not populated; however, the content object metadata is cleared out of the database. Customers can require that this time stamp is populated for auditing purposes that state that content has been removed from the system.

CSCth02785

Symptom:

Contents on the VVIM do not completely match the contents on the Vault.

CSCti38354

Symptom:

During a new cache server addition, the remoteservers file is not updated with all of the vault servers.

CSCti73302

Symptom:

In 2.3.2 b20 when the customer executes a db_shutdown command avsdb generates a core file.

CSCtk68063

Symptom:

A local server (server A) has fewer contents than the master Vault/SSV (server B).

CSCtk69852

Symptom:

During adding Vault 10.74.115.134, all other servers are online, also online status check for other servers are passed, after cdsconfig returned and said that DB full sync completed and new server is added into system. Check other servers' .arroyorc, only Vault 10.74.115.124 and Vault 10.74.115.136's .arroyorc are updated, all other servers' .arroyorc are not updated.

CSCtk94473

Symptom:

When dynamically adding a new server 10.74.115.134, after it is complete all other servers' .arroyorc are updated correctly, but cdsconfig prompted that new adding failed from avsdb log. Database was shutdown after failure happen.

File Systems

CSCth87521

Symptom:

2.2.2 Multiple copies of Goids were created causing KDB m_lastWriteStart.

FSI

CSCti91796

Symptom:

213 in FSI response is assumed successful. There is an issue getting the file size from the ftp server as part of the ingest process. The ftp server is returning a 550 response for the SIZE request. There is also a source code issue. This response is parsed for a 213 response code, which is the success code for the ftp SIZE response. If you notice the 550 response contains the 213 characters:

Installation

CSCth87710

Symptom:

The /arroyo partition on CDE250-2S6 streamer has only 8GB disk space. It is much smaller than the /arroyo partition on GEN2 streamer which has 35GB space. With 8GB space it would be filled out by the log files within one day if we keep running a large streaming load, like 7K-10K streams.

CSCtj23608

Symptom:

Installation fails during RPM installation with error message: DB_VERSION_MISMATCH.

ISA

CSCth34073

Symptom:

The session plays from the beginning if the ISA stream setup fails during the session resume.

Conditions:

The ISA Stream Setup fails and the API Stream.status() return 0 npt to the OpenStream.

CSCth63889

Symptom:

If the CDSM gets stuck, it cannot persist the trick-mode data for the sessions, and the data is lost if the CDSM or avsdb is restarted.

If the trick-mode events are not dispatched to OpenStream, OpenStream goes into reconciliation mode. ISA Stream Service returns an empty list of the sessions, because it does not have the trick-mode data. OpenStream keeps calling StreamService.getTrickmodeEventsList() every 10 seconds.

Conditions:

The trick-mode data is lost some how.

CSCtg09554

Symptom:

Delete an IP subnet from the VHO CDSM and look into the VHO OpenStream for returnRoutings. The returnRoutings still shows the deleted IP Subnet. This does not impact any normal functionality though.

Conditions:

Deleting an IP Subnet.

CSCth98418

Symptom:

ISA ContentStore restarts because of network issues. If the network issues occur then it could cause multi-master. If multi-master occurs then content store or isa service monitor application might not get the reply back for its requests and could wait forever and so script times out and restart both monitor application and the content store.

CSCti21413

Symptom:

If the CDS server is unable to get to the OpenStream for registration, then it is quite hard to figure out what underneath causing the issue.

CSCtk01542

Symptom:

If the Content is already deleted from the Content table, then the MSV record for that content from the pending delete table should be deleted. The script also archive the log file.

CSCtk56053

Symptom:

Provision 2 packages with the same content name. The ContentStore will not throw the duplicate exception if 2nd content is getting provisioned.

Memory

CSCti68313

Symptom:

When a new vault is added to an array, the scope of other systems that it has visibility to is larger than can fit in an 8K buffer and the memory allocation fails. This causes an overflow condition and the vault crashes.

Monitoring

CSCtj83961

Symptom:

The test bed works on 2.3.3 b16 rtsp environment. If it ingest live recording and play live stream using this content in some trick speed, after the recording finish, the content copies on Completed Ingest page shows some goids are partially on the streamer.

CSCtk58766

Symptom:

When System Temperature reported by SDT exceeds the high limit, VVIM reports System Temperature incorrectly in the Server Vitals page.

High limit becomes concatenated with the label, e.g., "System Temp 50C/122F", current value is printed as 0C/32F and actual current value (e.g. ,"51C/123F") is incorrectly reported as maximum limit.

Network

CSCtj86813

Symptom:

Each system (vault, cache node, or streamer) keeps track of all the other servers that is within its scope of visibility. That allows for each of the systems to communicate, transfer content, provide management communications, provide fault tolerance and resiliency and maintain general health of the system. A table is kept in memory called the remote servers table and is used for keeping track of all other servers that are visible to each server. This remote servers table has a limit of 125 entries. A check was in the code to not allow more than 125 entries so that the table would not overflow and cause memory corruption. An error in the logic of how to handle this condition did not allow for appropriate operation of the system and therefore it went into kdb. Since this same condition exists in the code that is running on each of the vaults, when the 126th server was brought up it was visible to all the other vaults, therefore they all executed the invalid logic in the code causing each of them to go to kdb.

CSCtj87648

Symptom:

When a cache server has two vault groups that it could fill from and one of the vault groups has all the vaults offline so that they will not accept ingests, this triggers a situation where none of those vaults in that group will be selected to fulfill the fill request. This creates a situation where most if not all fill request made by the cache server will go to the one vault group and the fill capacity is not balanced.

CSCtk11488

Symptom:

Duplication of content across caching servers.

Other

CSCso89545

Symptom:

Performance test for 60-channels keeps going working with BMS. 3 core dumps are found when creating and publishing ADI packages. Confirmed by BMS, at that time, OpenStream servers was rebooting.

CSCtg97814

Symptom:

/arroyo/test/RemoteServers file on the streamer is populated with all the cache servers IPs.

CSCth40761

Symptom:

After rebooting a Vault, the Trickle_Down file is missing.

CSCth83720

Symptom:

Error in CDSM /arroyo/www/logs/error_log:

/arroyo/www/htdocs/scripts' not found or unable to stat
 
   

CSCti05893

Symptom:

While running cdsconfig on a streamer in a CDSM Domain ( Split Domain using HTTP as cache fill protocol), it doesn't accept a Server ID greater than 1000.

CSCtj21349

Symptom:

Ingests on the platform will not start.

CSCtk09042

Symptom:

During the ingest process where Widevine encryption is utilized, AIM issues a PackageNotifyRequest. Widevine provides a HTTP response that utilizes chunked encoding for the XML payload. On occasion, the next chunk is not immediately in the socket and AIM is failing the ingest request. AIM should be more robust and allow for the situation where the next chunk is not immediately available on the socket.

RTSP

CSCtg60835

Symptom:

RTSP server issues a 551 Option Not Supported response for requests that don't include the .decimal_npts require value.

CSCti83038

Symptom:

RTSP requests are not being processed and/or are being processed very slowly (long latency between request and response).

SNMP

CSCth34909

Symptom:

When performing an SNMP GET or GETNEXT on some ifXTable objects, e.g., ifHighSpeed.4, SNMPD crashes almost randomly. SNMP walk on this table produces similar results.

statsd

CSCti83139

Symptom:

In ISA environments where Vault redundancy is disabled and National and Local Vaults are configured, it is possible that the Setup Stream Group does not have the servers from the other Control-only Stream Groups.

Conditions:

ISA environments, Vault redundancy disabled, National and Local Vaults are configured.

CSCtk69250

Symptom:

In customer there are four streamer groups: ZZ-Streamer-Group1, ZZ-Streamer-Group2, ZZ-Streamer-Group3, and XX-Streamer-Group1.

ZZ-Streamer-Group1 is the setup streamer group. Normally, remoteservers file from setup streamer group should have all other streamers; remoteservers file from non setup streamers should only have peer streamers in the same group and streamers in the setup groups.

Right now, streamer in other three streamer groups have wrong remoteservers file, they include streamer from other non setup streamer groups.

CSCtk82042

Symptom:

Allows for new (multiple) power supply fan, temp. and failure monitoring.

Streamer

CSCtg51516

Symptom:

Content that has been pre-encrypted for Motorola STBs sometimes does not play the first time. Subsequent play attempts sometimes succeed.

CSCtg78367

Symptom:

Streamer in kdb state with back trace showing SC::Stream::setStreamInfo+0x619 as the location of the page fault.

CSCti79602

Symptom:

If RTSP is having issues and we move the master then we may crash the backup.

Utilities

CSCtg25603

Symptom:

On avs_clist on 1.5.1.5.4, the clone list has problems with the goid length. Also. the cloned vault can delete a remote goid.

Vault

CSCtg29316

Symptom:

This fix is to prevent a malformed vaultcloner file from causing the cloner to delete content on the donor.

CSCth97509

Symptom:

Evaluator stuck in disabled Mode. In the protocoltiming.log, the following warning is seen:

WARNING: Evaluators are disabled.
 
   

CSCtj00762

Symptom:

Vault kdb when more than 20 copies are created:

Unable to handle kernel paging request at 000000a087d65de0
 
   

CSCtj61822

Symptom:

Evaluator is stuck in disable mode on several vaults.

CSCtk69771

Symptom:

Vault repeatedly enters kdb at "../rom/ObjectUpgrader.cpp", line 1234.

CSCtk97769

Symptom:

One vault hit ASSERT "LOM::ObjectLocalMirrorAgent invalid remove\n" when concurrently ingest different rate VOD contents, include MPEG SD, MPEG HD, H.264 SD, H.264 HD, non-SD.

Video Quality

CSCtg36215

Symptom:

Macroblocking for up to a second after a transition from FF or REW to 1x Play.

CSCti79546

Symptom:

Content ftpout size mismatched the original size.

CSCti19189

Symptom:

Occurs during tick mode when fast-forward is selected the IMG timer shows 2:10, then when selecting play, the IMG timer shows 2:25 and the video jumps. Other times if fast-forward has been selected and then play is selected, a five-second jump can be seen.

CSCtj24141

Symptom:

If OLES encryption is enabled in an IP VOD environment, and jitter is measured at stream output from CDS TV, the jitter will exceed the MPEG-2 specs.

CSCtj59696

Symptom:

Content fails ingest.

CSCtk32629

Symptom:

Content won't ingest.

Upgrading to Release 2.3.3

For new installation and upgrade procedures, see the Cisco TV CDS 2.3 Installation, Upgrade, and Maintenance Guide.

If custom modifications have been made to the root crontab (for example, in ISA environments the MSV script to delete orphaned content), create the /etc/cds_noCrontabOverwrite file before upgrading the TV CDS software in order to avoid the custom crontab being overwritten. To create this file, enter the following command:

touch /etc/cds_noCrontabOverwrite
 
   

When upgrading to Release 2.3.3 or installing Release 2.3.3, you must run the cdsconfig script in order to disable the Redirect Server for RTSP Deployments. The following prompt has been added to the cdsconfig script:

Do you want to enable Redirect Server ? (yes/no) [y]
 
   

Select no if you need to disable the Redirect Server. It is not required to run cdsconfig for ISA/RTSP Shim Deployments.


Note Release 2.3.3 supports upgrades from Release 2.0.x and Release 2.1.x. If your CDS is running an older version, you need to upgrade to Release 2.0.x or 2.1.x, before upgrading to Release 2.3.

If downgrading the VVIM or CDSM from Release 2.3.3 to Release 2.2.1, you must remove the /arroyo/www/htdocs/cdsm directory before downgrading the software.



Note The following TV CDS software releases can coexist during an upgrade process, which allows for long upgrade windows, if necessary:

For Vaults in a VVI in an ISA environment with SCS, the Release 2.2.1 ES8 software can coexist with the Release 2.3.3 software.

For Caching Nodes in a VVI in an ISA environment with SCS, the Release 2.2.1 ES9 software can coexist with the Release 2.3.3 software.

For Streamers in a VVI in an ISA environment with SCS, the Release 2.2.1 ES5 software can coexist with the Release 2.3.3 software.

For VVI in an RTSP environment with HTTP Streamers, the Release 2.3.2 ES3 software can coexist with the Release 2.3.3 software.

For CDS (non-VVI) in both ISA and RTSP environments, the Release 2.0.x, 2.1.x, and 2.3.2 software can coexist with the Release 2.3.3 software.

The above TV CDS software releases can only coexist during an upgrade process, and only if the upgrade sequences described in the Cisco TV CDS 2.3 Installation, Upgrade, and Maintenance Guide are followed.

Different TV CDS software releases are not supported in the long term on an ongoing basis.


Updating the Device Drivers for new SSDs

The SSD driver update is for new SSDs and to address an issue with older SSDs. For more information, see the "New Features" section.

If a CDE requires an SSD replacement, the device driver must be updated. The device driver must be updated before replacing the SSD. If a new CDE is being added to the CDS, after upgrading the TV CDS software, the device driver must be updated. New SSDs are not recognized with the proper attributes until the device driver is updated.


Note The new driver is 100 percent backward-compatible with older, pre-existing SSD drives. We recommend that a proactive upgrade of the SSD driver is a good best-practice.


To view a list of TV CDS software releases this device driver applies to see the defect, CSCtr29064, in the Bug Tool Kit ((http://tools.cisco.com/Support/BugToolKit/).


Note When using the cdsinstall script to install the software image file, install the driver package after the cdsinstall script is complete.


To install the device driver update package, do the following:


Step 1 Download the cdstv_x.y.z_CSCtr29064.bin driver update package to the /root directory of the CDE. The filename is cdstv_x.y.z_CSCtr29064.bin,where x.y.z denotes the software release.


Note The driver update package includes support for all applicable software releases, so any cdstv_x.y.z_CSCtr29064.bin file installs correctly and can be used to update the drivers for any applicable release.


Following is the md5 checksum for the cdstv_x.y.z_CSCtr29064.bin driver update package:

89684e2094d841b236ba1d83e958df9b cdstv_x.y.z_CSCtr29064.bin 
 
   

The driver update package is a self-extracting, self-installing package.

Step 2 Change the permissions on the CDE. As user root, enter the following command:

# chmod 755 /root/cdstv_x.y.z_CSCtr29064.bin 
 
   

Step 3 Install the device driver update. As user root user enter the following commands:

# cd /root
# ./cdstv_x.y.z_CSCtr29064.bin 
 
   

The package installs the device driver.

Step 4 To verify that the package has been installed on the CDE, view the /etc/cisco-version-history file, and verify that the string "cdstv_CSCtr29064" has been added to this version-history file.

Step 5 To verify that the driver has been updated and will be used by the CDS server, enter the following commands:

cd /lib/module/`uname -r` 
/sbin/modinfo -F version csd.ko cdd_sys.ko
 
   

The updated driver files report the same version for both driver files in the format of x.y.z.1, where x.y.z is the version of the TV CDS software (for example, 2.3.3.1).

Step 6 Reboot the CDS server.

# reboot
 
   

Note Rebooting a Vault server does not interrupt stream services, but causes current ingests to fail. If your CDS does not have stream failover, rebooting a Streamer without offloading it interrupts all stream services. If possible, you should perform functions that require a system restart during times when the least number of users are actively connected to your system.



A new SSD drive is a Generation 2 front-mounted SSD. The new SSD model is identified by the title "Intel SSD 320 Series" on the label and has the model number: SSDSA2BW160G3. For more information, see Field Notice 63438 at: http://www.cisco.com/en/US/products/ps7126/prod_field_notices_list.html.

The following command provides information on the external SSDs that are the new drive model:

# find /sys/devices -name "model" | xargs grep SSDSA2BW16
/sys/devices/pci0000:80/0000:80:05.0/0000:87:00.0/host3/port-3:7/end_device-3:7/target3:0:7/3:0:7:0/model:INTEL SSDSA2BW16
/sys/devices/pci0000:80/0000:80:05.0/0000:87:00.0/host3/port-3:6/end_device-3:6/target3:0:6/3:0:6:0/model:INTEL SSDSA2BW16
/sys/devices/pci0000:80/0000:80:05.0/0000:87:00.0/host3/port-3:5/end_device-3:5/target3:0:5/3:0:5:0/model:INTEL SSDSA2BW16
/sys/devices/pci0000:80/0000:80:05.0/0000:87:00.0/host3/port-3:4/end_device-3:4/target3:0:4/3:0:4:0/model:INTEL SSDSA2BW16
/sys/devices/pci0000:80/0000:80:05.0/0000:87:00.0/host3/port-3:3/end_device-3:3/target3:0:3/3:0:3:0/model:INTEL SSDSA2BW16
 
   

To view all SSD models currently supported by the Cisco TV CDS software, enter the following command:

# find /sys/devices -name "model" | xargs grep SSDSA
 
   

The older SSD model is INTEL SSDSA2M160.

Rolling Back the SSD Driver Update

If for any reason you need to revert to the older SSD driver, with the TV CDS software running on the CDS server, do the following:


Step 1 Change to the drivers directory.

cd /lib/module/`uname -r` 
 
   

Step 2 To save the updated driver (optional), enter the following command:

cp csd.ko csd.ko.update 
cp cdd_sys.ko cdd_sys.ko.update
 
   

Step 3 Rename the backed up driver files to the driver filenames. The "previous_version" is the version number of the driver update.

cp csd.ko_bkup_<previous_version> csd.ko
cp cdd_sys.ko_bkup_<previous_version> cdd_sys.ko
 
   

Step 4 Verify the driver versions by using the following command:

/sbin/modinfo -F version csd.ko cdd_sys.ko
 
   

The drivers before the driver update have a version in the format of a.b.c and the two drivers do not necessarily have the same version number.

Step 5 Reboot the CDS server.


Related Documentation

Refer to the following documents for additional information about the Cisco TV CDS 2.3:

Cisco TV CDS 2.3 ISA Software Configuration Guide

http://www.cisco.com/en/US/docs/video/cds/cda/tv/2_3/configuration/isa_guide/tv_cds_2_3_isa_config.html

Cisco TV CDS 2.3 RTSP Software Configuration Guide

http://www.cisco.com/en/US/docs/video/cds/cda/tv/2_3/configuration/rtsp_guide/tv_cds_2_3_rtsp_config.html

Cisco TV CDS 2.1-2.3 API Guide

http://www.cisco.com/en/US/docs/video/cds/cda/tv/2_1/developer_guide/tv_cds_2_1_apiguide.html

Cisco TV CDS 2.3 Installation, Upgrade, and Maintenance Guide

http://www.cisco.com/en/US/docs/video/cds/cda/tv/2_3/installation_guide/tv_cds_2_3_maint_guide.html

Cisco Content Delivery System 2.x Documentation Roadmap

http://www.cisco.com/en/US/docs/video/cds/overview/CDS_Roadmap.html

Cisco Content Delivery Engine 205/220/250/420 Hardware Installation Guide

http://www.cisco.com/en/US/docs/video/cds/cde/cde205_220_420/installation/guide/cde205_220_420_hig.html

Cisco Content Delivery Engine 110 Hardware Installation Guide

http://www.cisco.com/en/US/docs/video/cds/cde/cde110/installation/guide/cde110_install.html

Cisco Content Delivery Engine 100/200/300/400 Hardware Installation Guide

http://www.cisco.com/en/US/docs/video/cds/cde/installation/guide/CDE_Install_Book.html

Regulatory Compliance and Safety Information for Cisco Content Delivery Engines

http://www.cisco.com/en/US/docs/video/cds/cde/regulatory/compliance/CDE_RCSI.html

The entire CDS software documentation suite is available on Cisco.com at:

http://www.cisco.com/en/US/products/ps7127/tsd_products_support_series_home.html

The entire CDS hardware documentation suite is available on Cisco.com at:

http://www.cisco.com/en/US/products/ps7126/tsd_products_support_series_home.html

Obtaining Documentation and Submitting a Service Request

For information on obtaining documentation, submitting a service request, and gathering additional information, see the monthly What's New in Cisco Product Documentation, which also lists all new and revised Cisco technical documentation, at:

http://www.cisco.com/en/US/docs/general/whatsnew/whatsnew.html

Subscribe to the What's New in Cisco Product Documentation as a Really Simple Syndication (RSS) feed and set content to be delivered directly to your desktop using a reader application. The RSS feeds are a free service and Cisco currently supports RSS version 2.0.