Lotus Domino 6.5.4 Fix Pack 1

Finalmente disponibile il Fix Pack 1 della R 6.5.4. Ecco l'elenco di funzionalità aggiunte e/o bug risolti

Administration	ASCR68ZRZ6	SPR# ASCR68ZRZ6 - When the Administration Process checks disk space for a new replica, it might match the wrong physical directory statistics and thus the wrong disk when deciding which disk a new replica should reside on. For example, if the 'mail' directory is on a different drive from the 'domino data' directory, Adminp might choose the 'domino data' directory (back one directory from mail). It would then use those drive statistics (the wrong drive) when testing for enough disk space. This was based on physical drive layout on the machine or drive search data. With this fix, Adminp will match only the correct physical drive statistics. This regression was introduced in 6.0.4. Refer to SPR# RCAR6DDLY4 for a similar problem where the statistics used for the drive that will hold the new replica were formatted in scientific notation and could not be correctly converted by Adminp.
Agents	HSAO6A3NZ3	SPR# HSAO6A3NZ3 - Fixed an HTTP crash on NqueryAccessRoles. An Agent contained code that returned the roles of the user. The crash occurred if the 'user' had no roles specified.
Client	TTRT69EUV5	SPR# TTRT69EUV5 - Fixed a crash during the initialization of the spell check used by DWA.
Compact	WSCN67FJT2	SPR# WSCN67FJT2 - This fix allows a bucket position in a bucket descriptor to be zero when freeing buckets in compact. Without this change, the database will always be marked corrupt when compacted to reclaim space.
Database	MIAS6AHQ44	SPR# MIAS6AHQ44 - The system may terminate abnormally and crash while doing a name lookup. This problem occurs if the Directory Assistance ordered replica list changes while the lookup is done. The fix realigns a pointer which can change when a semaphore is released.
Database	TMIZ63JAPB	SPR# TMIZ63JAPB - Fixed a problem where the router was unable to resolve a user in the address book. Namelookup from the Notes client worked fine for this user, but mail was not delivered since the mechanism the router used to resolve the name used a different search mechanism. This search mechanism is now fixed and the names should resolve to allow mail delivery.
GTR	THTO64DSZ2	SPR# THTO64DSZ2 - Fixed a crash due to a large full-text search result set.
Java	GVAS66LMXY	SPR# GVAS66LMXY - HTTP crash caused by concurrently running a Java agent that uses the DbDirectory.OpenDatabaseByReplicaID method.
LotusScript	LVAE68YPNE	SPR# LVAE68YPNE - Fixed a "PANIC: LookupHandle: null handle" error. The problem was that a pointer field was not being cleared when a buffer had been freed. This pointer was left untouched, and resulted in a null handle being passed, causing the panic.
MIME	THTO68YUBB	SPR# THTO68YUBB - Fixed a crash with a large rfc822 header in a nested message. The extra large 'To' header overflowed the allocated memory, which may or may not cause a crash, depending on whether the process owned the memory after the overflowed buffer.
MIME	TMIZ6BSDLT	SPR# TMIZ6BSDLT - Fixed a problem which caused spaces to be dropped from the subject when rfc2047 encoding was used. This regression was introduced in 6.5.4.
MIME to CD conversion	MDLS54ZTJT	SPR# MDLS54ZTJT - When MIME_CONVERT_HTML_TO_ATTACHMENT=1 was set, and an incoming message contained HTML parts, which are quoted=printable encoded, the quoted printable parts do not get decoded, and R4.6x clients see the encoding when they look at the c.htm attachment.
Miscellaneous	ESAO62PMKP	SPR# ESAO62PMKP - A NULL value can be potentially returned when calling GetContentItemValue, leading to a crash. The fix is to exit the function when we see a NULL.
Networking & Dialup	CCAY64CR5S	SPR# CCAY64CR5S - Fixed an intermittent "Network operation did not complete in a reasonable amount of time; please retry." error case. Users could successfully retry the operation after the error message was displayed. Customers have reported that this problem sometimes occurred after they were inactive for a longer period of time (30 minutes or more) and they went to open a database on a server.
Networking & Dialup	PMIA65JSMC	SPR# PMIA65JSMC - The server based web retriever was sending hundreds of async notifications to the client to keep the client informed of web retrieval progress. The IOCP mechanism cannot handle this load of async IO. This change stops the web retriever from sending the notifications.
Networking & Dialup	RBEO639H7Q	SPR# RBEO639H7Q - Better recovery from transient low memory conditions. The port driver code which mapped port and platform specific network errors into common Notes/Domino error codes assumed that all network data structures were available. If the network operation failed due to low memory conditions, then some data structures were not available and a crash resulted. Fix is to check that all data structures are available before they are accessed.
Networking & Dialup	VDES68DF3J	SPR# VDES68DF3J - Better recovery from transient low memory conditions. Domino internally uses a signal handling mechanism to process certain events. The routine which returns the signal handler for a specific signal was not checking for memory allocation failures and was crashing. The fix is to check for memory allocation failures and return no handler rather than crashing.
Performance	JPAS6ALJQE	SPR# JPAS6ALJQE - Fixed a problem where the NIF pool ran out of memory after the NIF Global Pool grew to approx 128 megs. Users were not able to connect and received the error message "Insufficient memory - index pool is full".
Performance	RGET67JUVB	SPR# RGET67JUVB - Performance issues related to long held locks (Lock Manager) on clubusy.nsf, as seen in the console.log. These long held locks occur hourly (to within seconds) and typically cause 3-5 minute delays. With this fix, we changed when SchedMgr gathers stats to improve server performance. The logic is no longer hourly and only occurs at startup and during the 2AM revalidation phase. A workaround to this performance problem would be to disable hourly reporting of Schedule Manager statistics (calendar.* stats) with the Notes.ini, "Schedule_No_Calcstats=1". This parameter requires the restart of Sched only. Although the calendar.* stats won't be cleared from the stats list until the entire server is rebooted, these stats will not be updated once SchedMgr has been restarted.
Router	JBAA6BALX7	SPR# JBAA6BALX7 - This fix prevents an infinite loop when the fallback ini (RouterFallbackNonTLS=1) is set and the server is not responding. This regression was introduced in 6.5.4.
Router	KSPR699NBP	SPR# KSPR699NBP - Fixed a potential security issue.
Sametime	DFIO664NSP	SPR# DFIO664NSP - Better recovery from transient low memory conditions, when the server was responding to an NRPC request that was bad.
Security	MKIN67MQVW	SPR# MKIN67MQVW - Fixed a memory leak during SSL handshake. This leak will not be seen in memory dumps, but customers might see HTTP grow in virtual memory up to the system crash.
Security	MKIN693QUT	SPR# MKIN693QUT - Fixed a memory leak in the stash file during SSL handshake. This leak will not be seen in memory dumps but customers might see HTTP grow in virtual memory up to the system crash.
Security	MKIN693QX4	SPR# MKIN693QX4 - Fixed a memory leak during X.509 certificate verification.
Security	PMIA66JJWK	SPR# PMIA66JJWK - Better recovery from transient low memory conditions. During the authentication phase of a connection to a Domino server, if a memory allocation failure occurred, a pointer to a data structure remained null. It was later used and caused a crash. This fix is to check if the pointer is null before using it.
Server	ATHS67US6Q	SPR# ATHS67US6Q - Fixed a crash in ConvertZoneToText(), if we fail to load the TimeZoneTable. This error is reported in the log file.
Server	HLEZ65HMWM	SPR# HLEZ65HMWM - Prior to this fix, database corruption in the Hash Table caused a Server Crash when fixup was run. This was an intermittent issue that would only occur with a database that had this type of corruption. There were no error messages.
Server	JNON6B2QWG	SPR# JNON6B2QWG - This fix prevents an intermittent session hang on Solaris. A workaround to this problem is to reboot the Domino server to free the session.
Server	JPAI676KV5	SPR# JPAI676KV5 - Fixed a crash during Fixup. This problem occurs when Fixup takes a long time to run and overlaps the Agent Manager cache refresh.
Server	KLME63GSCS	SPR# KLME63GSCS - Fixed a panic during FixUp trying to process a DBlist which was not present. The code was not checking the returned value from dbListLookupIndex. In this case it is returning -1, which means there is no DbList. With this fix, a -1 would prevent the lookup from trying to process the DbList.
Server	MOAH683PCQ	SPR# MOAH683PCQ - Fixed an intermittent hang when the 'show stats' command was issued. The log.nsf file was not updated but users were able to connect to the server without a problem. A workaround to this problem would be to restart the server.
Server	RBEO62DGEF	SPR# RBEO62DGEF - Fix for bad error path which could cause a crash in NetAbortCheck.
Server	SKAI67GEW6	SPR# SKAI67GEW6 - This fix prevents a server from freezing after the "Drop" command is issued to drop all users who have the given database open.
Server	TCHL6A6QPZ	SPR# TCHL6A6QPZ - This fixes a potential problem when indexing XLS attachment files using the Keyview filters. This is not a problem with all XLS files, and does not affect the Domino server. It does prevent certain databases from being indexed if the bad attachments are included.
Server	TPOK6CKQW4	SPR# TPOK6CKQW4 - This fix prevents an intermittent server crash during HTTP refresh. This regression was introduced in 6.5.4. The problem appeared because an array of pointers was not properly initialized on the iSeries platform. As a result, if the memory allocated to hold the configuration pointers happened to contain an old pointer, a code change in 6.5.4 would attempt to use the pointer in an incorrect way resulting in either MCH3601 or MCH3602 errors. A workaround to this problem would be to set the HTTPConfigRefreshTimeout=1440 parameter in the Notes.ini. Please note that the timeout is in minutes, so 1440 is 24 hours. Basically, you should adjust the time to be larger than the expected uptime of the Domino Server to avoid a crash. Also, as part of this workaround, avoid issuing the "tell http refresh" command at the Domino Console or via any program document or Agent.
Serviceability	AEDR65ZNGB	SPR# AEDR65ZNGB - This corrects the timeout code in memcheck on z/OS. If memcheck is stuck, the normal timeout will now work and memcheck will return. This issue only impacts zSeries, other plaforms do not have this issue.
Serviceability	GCHN69DJKQ	SPR# GCHN69DJKQ - This fix will help service issues where the console traceback has not been able to show the functions involved and/or their parameters.
Serviceability	JBAA6CYSDF	SPR# JBAA6CYSDF - On Windows, NSD execution time ignores the Cleanup Script / NSD Maximum Execution Time setting in the Domino Server Document. The default value is 300 seconds but would run for up to 1 hour if it did not complete within the Max Execution time limit specified.
Serviceability	RGET65QRMD	SPR# RGET65QRMD - Eliminated flaws in how NSD validated stack frames during stack annotation, one of which had been the cause for the infinite loop encountered by several customers. When the problem occurs, NSD goes into an infinite loop reading stack frames and generating output in an internal buffer. What the customer sees is that NSD runs indefinitely, taking CPU and virtual memory until manually terminated. There are no other obvious symptoms, except that the resulting NSD output log only contains the initial header section (the output up to but not including stack trace output). This regression was introduced in 6.5.2.
Statistics	AEDR6B7JVT	SPR# AEDR6B7JVT - On zSeries, gathering SMF task statistics used a lot of resources and was changed to run every minute in error. When there are a lot of tasks running the statistics might not complete within a minute. This fix only gathers SMF task statistics every 20 minutes. This regression was introduced in 6.5.4.
Web Server	HSAO6BNL6Y	SPR# HSAO6BNL6Y - Fixed a potential security issue.
Web Server	KEMG6ATST6	SPR# KEMG6ATST6 - It is now possible to disable the web server from processing passthru html. This is a server wide option. The option is controlled with the configuration variable: DominoDisablePassThruHTML 0 = (the default) enables passthru html 1 = disable passthru delimited with square brackets 2 = disable all passthru html
Web Server	KSPR66USSU	SPR# KSPR66USSU - Fixed a potential security vulnerability.
Web Server	NORK65PNF3	SPR# NORK65PNF3 - When moving from Domino 5.x to 6.x, the HTTP thread management model was changed. With the 5.x model, each thread managed one request at a time. In 6.x, the model was changed so that each thread could have a queue of requests. With this change in 6.5.4 FP1, and 6.5.5, we are allowing the option to change how threads process HTTP connections. The original R5.x model or R6.x queue of requests model can be selected. Use the following Notes.ini settings for these changes: HTTPQueueMethod=0 - This setting is the default, and represents no change in queueing from the R6.0 and above. The accept thread will evenly distributed network connections to all worker threads using a round robin method. Connections are assigned to a specific thread. HTTPQueueMethod=1 - This setting will cause the accept thread to find the worker thread that has the least number of network connections assigned to it and assign the new network connection to that worker thread. It is still possible that a new network connection may be assigned to a thread that is in the process of processing a long running request. It is recommended but not required that persistent connections are disabled when using this option to get the maximum effect. This will limit the possible wait time if a new network connection is queued to a thread that already is busy with other connections(s) HTTPQueueMethod=2 - This setting will cause the accept thread to put the incoming network connections on a queue from with the worker threads will pull from. This is the same model as R5. It is also recommend but not required that persistent connections be disabled when using this option to get the maximum effect. In general the best method to use will depend on the nature of applications running on the server. If the mix of URL requests presented to the server run very quickly then option 0 or 1 will be the best option. Option 1 performs a little better then option 0 but at a slightly higher CPU cost so if URL requests are CPU bound then option 1 may actually slow overall through put of the server. The slowest option is option 2 with regard to overall server throughput, however if the mix of URL requests include very long running requests such as large uploads/downloads or URLs that run application code the this may be the best option for the customer.