ctdb-devel-1.0.89-2.4>t  DH`pJ{ָ/=„@d;:o-ʿ>``]DiQPnY&{Ǫ3em)PϦH4D lQC1!SeV{r*_C|+NNSDjMu@OxDT4t^`]iJ'Gu゘43$b4m\ x9Q8n5T<\2k8g; @]\9[O􃢼`H-W06db493a4d9de1eecd84bf2ab89e19534a393f4d͐J{ָ/=„3Jkyq4!8suEK#7rmᏄ'(#+h%Dpփʙ㞥!3ZwP WKkjwzwzi6!s'ؓh~OelqL3 0IមtxUEv Xe(O`%.R։Zp,Q.uhFr",`UAIr.zqzYHSZ~q4|T>5?d   Y :QW`l r x   0<\h(8H9H:vHFkGHIXY\]^b c}deflzCctdb-devel1.0.892.4Libraries and Header Files to Develop Programs with ctdb supportThis package contains the static libraries and header files needed to develop programs which make use of ctdb.J{build249openSUSE 11.2openSUSEGNU GPL version 3http://bugs.opensuse.orgDevelopment/Libraries/C and C++http://ctdb.samba.org/linuxx86_64UJ{J{J{48f4cd93079abb5d0438dad79de6bb875183e3b83ecb034b64070193761b62d010e804e4973e169a74fcf9953c38189crootrootrootrootrootrootctdb-1.0.89-2.4.src.rpmctdb-develctdb-devel(x86-64)   rpmlib(PayloadFilesHavePrefix)rpmlib(CompressedFileNames)rpmlib(PayloadIsLzma)4.0-13.0.4-14.4.6-14.7.1J@J67J67J/@J.NJ@J@J@IܑIII@I@I?@IV@Im@I@I@IpvIJ;@IBR@I6tI*@IHHrH,H@H@HO@HAHH)H)H5@HH@HHH4@H4@HwK@HuHsVHr@Hj@HRa@H=I@H;H5`@H-w@H(1@H$/dev/null for grep to enhance readability of the scripts from Michael Adam. + Update to the "ctdb recover" command. This command now block until the has completed. This makes it much easier to use in scripts and avoids the common workaround : ctdb recover ... loop while waiting for recovery completes ... continue ... + Add a CTDB_TIMEOUT variable. If set, this variable provides an automatic timeout for "ctdb ", similar to using -T + Set a unique errorcode for "ctdb " when it terminates due to a timeout so that scripts can distinguish between a hung command and what was just a failure. + Update "ctdb ban/unban" so that if the cluster is in recovery these commands blocks and waits until after recovery is complete before the perform the ban/unban operation. This is necessary since the recovery process can cause nodes to become automatically unbanned. + Update "ctdb ban/unban" to block until the recovery that will follow shortly after this command has completed. This makes it much easier to use in scripts and avoids the common workaround : ctdb ban/unban ... loop while waiting for recovery completes ... continue ... + Bugfix for the new flags handling in 1.0.67. Abort and restart monitoring if we failed to get proper nodemaps from a remote node instead of dereferencing a null pointer. + If ctdbd was explicitely started with the '--socket' argument, make ctdbd automatically set CTDB_SOCKET to the specified argument. This ensures that eventscripts spawned by the ctdb daemon will default to using the same socket and talk to the correct daemon. This primarily affects running multiple daemons on the same host and where you want each instance of ctdb daemons have their eventscripts talk to the "correct" daemon. + Update "ctdb ping" to return an error code if the ping fail so that it can be used in scripts. + Update to how to synchronize management of node flags across the cluster.- Version 1.0.67 + Add a document describing the recovery process. + Fix a bug in "ctdb setdebug" where it would refuse to set a negative debug level. + Print the list of literals for debug names if an invalid one was given to "ctdb setdebug" + Redesign how "ctdb reloadnodes" works and reduce the amont of tcp teardowns used during this event. + Make it possible to delete a public ip from all nodes at once using "ctdb delip -n all"- Version 1.0.66 + Allow to change the recmaster even when we are not frozen. + Remove two redundant SAMBA_CHECK variables from the sysconf example + After a node failure it can take very long before some lock operations ctdb needs to perform are allowed/works with gpfs again. Workaround this by treating a hang/timeout as success. + Dont override CTDB_BASE is fet in the shell already + Always send keepalive packets regardless of whether the link is idle or not. + Rewrite the disable/enable flag update logic to prevent a race between "ctdb disable/enable" and the recovery daemon when updating the flags to all nodes.- Version 1.0.65 + Update the sysconfig example: The default debug level is 2 (NOTICE) and not 0 (ERROR) + Add support for a CTDB_SOCKET environment variable for the ctdb command line tool. If set, this overrides the default socket the ctdb tool will use. + Add logging of high latency operations.- Version 1.0.64 + Add a context and a timed event so that once we have been in recovery for too long we drop all public addresses.- Version 1.0.63 + Remove logging of "periodic cleanup ..." in 50.samba + When we reload a nodes file, we must detect this and reload the file also in the recovery daemon before we try to dereference somethoung beyond the end of the nodes array.- Version 1.0.62 + Allow multiple eventscritps using the same prefix number. It is undefined which order scripts with the same prefix will execute in.- Version 1.0.61 + Use "route add -net" instead of "ip route add" when adding routes in 99.routing + lower the loglevel os several debug statements + check the status returned from ctdb_ctrl_get_tickles() before we try to print them out to the screen. + install a new eventscript 20.multipathd whoich can be used to monitor that multipath devices are healthy- Version 1.0.60 + Verify that nodes we try to ban/unban are reachable and print an error othervise. + Update the client and server sides of TAKEIP/RELEASEIP/GETPUBLICIPS and GETNODEMAP to fall back to the old style ipv4-only controls if the new ipv4/ipv6 controls fail. This allows an ipv4/v6 enabled ctdb daemon to interoperate with earlier ipv4-only versions of the daemons. + From Mathieu Parent : log debian systems log the package versions in ctdb diagnostics + From Mathieu Parent : specify logdir location for debian (this patch was later reversed) + From Michael Adams : allow # comments in nodes/public_addresses files- Version 1.0.59 + Updated "reloadnodes" logic. Instead of bouncing the entire tcp layer it is sufficient to just close and reopen all outgoing tcp connections. + New eventscript 99.routing which can be used to re-attach routes to public interfaces after a takeip event. (routes may be deleted by the kernel when we release an ip) + IDR tree fix from Jim Houston + Better handling of critical events if the local clock is suddenly changed forward by a lot. + Fix three slow memory leaks in the recovery daemon + New ctdb command : ctdb recmaster which prints the pnn of the recmaster + Onnode enhancements from Martin S : "healthy" and "connected" are now possible nodespecifiers + From Martin S : doc fixes + lowering some debug levels for some nonvital informational messages + Make the daemon daemon monitoring stronger and allow ctdbd to detect a hung recovery daemon. + From C Cowan : patches to compile ipv6 under AIX + zero out some structs to keep valgrind happy- Version 1.0.58 + revert the name change tcp_tcp_client back to tcp_control_tcp so samba can build. + Updates to the init script from Abhijith Das - Version 1.0.57 + initial support for IPv6- Version 1.0.56 + fix a memory leak in the recovery daemon.- Version 1.0.55 + Fix the releaseip message we seond to samba.- Version 1.0.54 + fix a looping error in the transaction code + provide a more detailed error code for persistent store errors so clients can make more intelligent choices on how to try to recover- Version 1.0.53 + Remove the reclock.pnn file it can cause gpfs to fail to umount + New transaction code- Version 1.0.52 + Send an explicit gratious arp when starting sending the tcp tickles. + When doing failover, issue a killtcp to non-NFS/non-CIFS clients so that they fail quickly. NFS and CIFS already fail and recover quickly. + Update the test scripts to handle CTRL-C to kill off the test.- Version 1.0.51 + Strip off the vlan tag from bond devices before we check in /proc if the interface is up or not. + Use testparm in the background in the scripts to allow probing that the shares do exist. + Fix a bug in the logging code to handle multiline entries better + Rename private elements from private to private_data- Version 1.0.50 + Dont assume that just because we can establish a TCP connection that we are actually talking to a functioning ctdb daemon. So dont mark the node as CONNECTED just because the tcp handshake was successful. + Dont try to set the recmaster to ourself during elections for those cases we know this will fail. To remove some annoying benign but scary looking entries from the log. + Bugfix for eventsystem for signal handling that could cause a node to hang.- Version 1.0.49 + Update the safe persistent update fix to work with unpatched samba servers.- Version 1.0.48 + Update the spec file. + Do not start new user-triggered eventscripts if we are already inside recovery mode. + Add two new controls to start/cancel a persistent update. A client such as samba can use these to tell ctdbd that it will soon be writing directly to the persistent database tdb file. So if samba is -9ed before it has eitehr done the persistent_store or canceled the operation, ctdb knows that the persistent databases 'may' be out of sync and therefore a full blown recovery is called for. + Add two new options : CTDB_SAMBA_SKIP_CONF_CHECK and CTDB_SAMBA_CHECK_PORTS that can be used to override what checks to do when monitoring samba health. We can no longer use the smbstatus, net or testparm commands to check if samba or its config is healthy since these commands may block indefinitely and thus can not be used in scripts.- Version 1.0.47 + Fix a double free bug where if a user striggered (ctdb eventscript) hung and while the timeout handler was being processed a new user triggered eventscript was started we would free state twice. + Rewrite of onnode and associated documentation.- Version 1.0.46 + Document both the LVS:cingle-ip-address and the REMOTE-NODE:wan-accelerator capabilities. + Add commands "ctdb pnn", "ctdb lvs", "ctdb lvsmaster". + LVS improvements. LVS is the single-ip-address mode for a ctdb cluster. + Fixes to supress rpmlint warnings + AXI compile fixes. + Change \s to [[:space:]] in some scripts. Not all RHEL5 packages come with a egrep that handles \s even same version but different arch. + Revert the change to NFS restart. CTDB should NOT attempt to restart failed services. + Rewrite of the waitpid() patch to use the eventsystem for handling signals.- Version 1.0.45 + Try to restart the nfs service if it has failed to respond 3 times in a row. + waitpid() can block if the child does not respond promptly to SIGTERM. ignore all SIGCHILD signals by setting SIGCHLD to SIG_DEF. get rid of all calls to waitpid(). + make handling of eventscripts hanging more liberal. only consider the script to have failed and making the node unhealthy IF the eventscript terminated wiht an error OR the eventscript hung 5 or more times in a row- Version 1.0.44 + Add a CTDB_VALGRIND option to /etc/sysconfig/ctdb to make it start ctdb under valgrind. Logs go to /var/log/ctdb_valgrind.PID + Add a hack to show the control opcode that caused uninitialized data in the valgrind output by encoding the opcode as the line number. + Initialize structures and allocated memory in various places in ctdb to make it valgrind-clean and remove all valgrind errors/warnings. + If/when we destroy a lockwait child, also make sure we cancel any pending transactions + If a transaction_commit fails, delete/cancel any pending transactions and return an error instead of calling ctdb_fatal() + When running ctdb under valgrind, make sure we run it with --nosetsched and also ensure that we do not use mem-mapped i/o when accessing the tdb's. + zero out ctdb->freeze_handle when we free/destroy a freeze-child. This prevent a heap corruption/ctdb crash bug that could trigger if the freeze child times out. + we dont need to explicitely thaw the databases from the recovery daemon since this is done implicitely when we restore the recovery mode back to normal. + track when we start and stop a recovery. Add the 'time it took to complete the recovery' to the 'ctdb uptime' output. Ensure by tracking the start/stop recovery timestamps that we do not check that the ip allocation is consistend from inside the recovery daemon while a different node (recovery master) is performing a recovery. This prevent a race that could cause a full recovery to trigger if the 'ctdb disable/enable' commands took very long. + The freeze child indicates to the master daemon that all databases are locked by writing data to the pipe shared with the master daemon. This write sometimes fail and thus the master daemon never notices that the databases are locked cvausing long timeouts and extra recoveries. Check that the write is successful and try the write again if it failed. + In each node, verify that the recmaster have the right node flags for us and force a push of our flags to the recmaster if wrong.- Version 1.0.43 + Updates and bugfixes to the specfile to keep rpmlint happy + Force a global flags update after each recovery event. + Verify that the recmaster agrees with our node flags and update the recmaster othervise. + When writing back to the parent from a freeze-child across the pipe, loop over the write in case the write failed with an error othervise the parent will never be notified tha the child has completed the operation. + Automatically thaw all databases when recmaster marks us as being in normal mode instead of recovery mode.- Version 1.0.42 + When event scripts have hung/timedout more than EventScriptBanCount times in a row the node will ban itself. + Many updates to persistent write tests and the test scripts.- Version 1.0.41 + Reactivate the safe writes to persistent databases and solve the locking issues. Locking issues are solved the only possible way, by using a child process to do the writes. Expensive and slow but... .- Version 1.0.40 + Read the samba sysconfig file from the 50.samba eventscript + Fix some emmory hierarchical bugs in the persistent write handling- Version 1.0.39 + Moved a CTDB_MANAGES_NFS, CTDB_MANAGES_ISCSI and CTDB_MANAGES_CSFTPD into /etc/sysconfig/ctdb + Lowered some debug messages to not fill the logfile with entries that normally occur in the default configuration.- Version 1.0.38 + Add machine readable output support to "ctdb getmonmode" + Lots of tweaks and enhancements if the event scripts are "slow" + Merge from tridge: an attempt to break the chicken-and-egg deadlock that net conf introduces if used from an eventscript. + Enhance tickles so we can tickle an ipv6 connection. + Start adding ipv6 support : create a new container to replace sockaddr_in. + Add a checksum routine for ipv6/tcp + When starting up ctdb, let the init script do a tdbdump on all persistent databases and verify that they are good (i.e. not corrupted). + Try to use "safe transactions" when writing to a persistent database that was opened with the TDB_NOSYNC flag. If we can get the transaction thats great, if we cant we have to write anyway since we cant block here.- Version 1.0.37 + When we shutdown ctdb we close the transport down before we run the "shutdown" eventscripts. If ctdb decides to send a packet to a remote node after we have shutdown the transport but before we have shutdown ctdbd itself this could lead to a SEGV instead of a clean shutdown. Fix. + When using the "exportfs" command to extract which NFS export directories to monitor, exportfs violates the "principle of least surprise" and sometimes report a single export line as two lines of text output causing the monitoring to fail.- Version 1.0.36 + fix a memory corruption bug that could cause the recovery daemon to crash. + fix a bug with distributing public ip addresses during recovery. If the node that is the recovery master did NOT use public addresses, then it assumed that no other node in the cluster used them either and thus skipped the entire step of reallocating public addresses.- Version 1.0.35 + During recovery, when we define the new set of lmasters (vnnmap) only consider those nodes that have the can-be-lmaster capability when we create the vnnmap. unless there are no nodes available which supports this capability in which case we allow the recmaster to become lmaster capable (temporarily). + Extend the async framework so that we can use paralell async calls to controls that return data. + If we do not have the "can be recmaster" capability, make sure we will lose any recmaster elections, unless there are no nodes available that have the capability, in which case we "take/win" the election anyway. + Close and reopen the reclock pnn file at regular intervals. Make it a non-fatal event if we occasionally fail to open/read/write to this file. + Monitor that the recovery daemon is still running from the main ctdb daemon and shutdown the main daemon when recovery daemon has terminated. + Add a "ctdb getcapabilities" command to read the capabilities off a node. + Define two new capabilities : can be recmaster and can be lmaster and default both capabilities to YES. + Log denied tcp connection attempts with DEBUG_ERR and not DEBUG_WARNING- Version 1.0.34 + When deleting a public ip from a node, try to migrate the ip to a different node first. + Change catdb to produce output similar to tdbdump + When adding a new public ip address, if this ip does not exist yet in the cluster, then grab the ip on the local node and activate it. + When a node disagrees with the recmaster on WHO is the recmaster, then mark that node as a recovery culprit so it will eventually become banned. + Make ctdb eventscript support the -n all argument.- Version 1.0.33 + Add facilities to include site local adaptations to the eventscript by /etc/ctdb/rc.local which will be read by all eventscripts. + Add a "ctdb version" command. + Secure the domain socket with proper permissions from Chris Cowan + Bugfixes for AIX from Chris Cowan- Version 1.0.32 + Add a control to have a node execute the eventscripts with arbitrary command line arguments. + Add a control "rddumpmemory" that will dump the talloc memory allocations for the recovery daemon. + Decorate the talloc memdump to produce better and easier memory leak tracking. + Update the RHEL5 iscsi tgtd scripts to allow one iscsi target for each public address. + Add two new controls "addip/delip" that can be used to add/remove public addresses to a node at runtime. After using these controls a "ctdb recover" if required to make the changes take. + Fix a couple of slow memory leaks.- Version 1.0.31 + Add back controls to disable/enable monitoring on a node. + Fix a memory leak where we used to attach CALL data to the ctdb structure when performing a local call. Memory which would be lost if the call was aborted. + Reduce the loglevel for the log output when someone connects to a non public ip address for samba. + Redo and optimize the vacuuming process to send only one control to each other node containing all records to be vacuumed instead of one control per node per record.- Version 1.0.30 + Update documentation cor new commands and tuneables + Add machinereadable output to the ip,uptime and getdebug commands + Add a moveip command to manually failover/failback public ips + Add NoIPFallback tuneable that prevents ip address failback + Use file locking inside the CFS as alternative to verify when other nodes Are connected/disconnected to be able to recover from split network + Add DisableWhenUnhealthy tunable + Add CTDB_START_AS_DISABLED sysconfig param + Add --start-as-disabled flag to ctdb + Add ability to monitor for OOM condition- Version 1.0.29 + Add a new command to make expansion of an existing cluster easier + Fix bug with references to freed objects in the ctdb structure + Propagate debuglevel changes to the recovery daemon + Merge patches to event scripts from Mathieu Parent : + MP: Simulate "service" on systems which do not provide this tool + MP: Set correct permissions for events.d/README + Add nice helper functions to start/stop nfs from the event scripts- Version 1.0.28 + Fix a problem where we tried to use ethtool on non-ethernet interfaces + Warn if the ipvsadm packege is missing when LVS is used + Dont use absolute pathnames in some of the event scripts + Fix for persistent tdbs growing inifinitely.- Version 1.0.27 + Add eventscript for iscsi- Version 1.0.26 + Fix crashbug in tdb transaction code- Version 1.0.25 + added async recovery code + make event scripts more portable + fixed ctdb dumpmemory + more efficient tdb allocation code + improved machine readable ctdb status output + added ctdb uptime- Version 1.0.24 + added syslog support + documentation updates- Version 1.0.23 + fixed a memory leak in the recoveryd + fixed a corruption bug in the new transaction code + fixed a case where an packet for a disconnected client could be processed + added http event script + updated documentation- Version 1.0.22 + auto-run vacuum and repack ops- Version 1.0.21 + added ctdb vacuum and ctdb repack code- Version 1.0.20 + new transaction based recovery code- Version 1.0.19 + fixed non-master bug + big speedup in recovery for large databases + lots of changes to improve tdb and ctdb for high churn databases- Version 1.0.18 + fixed crash bug in monitor_handler- Version 1.0.17 + fixed bugs related to ban/unban of nodes + fixed a race condition that could lead to monitoring being permanently disabled, which would lead to long recovery times + make deterministic IPs the default + fixed a bug related to continuous recovery + added a debugging option --node-ipbuild24 1256356799EF@1.0.89-2.41.0.89-2.4ctdb.hctdb_private.hctdb.pc/usr/include//usr/lib64/pkgconfig/-fmessage-length=0 -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector -funwind-tables -fasynchronous-unwind-tables -gobs://build.opensuse.org/openSUSE:11.2/standard/a095f67ccc5f7cba8e05dbb72c0eebbf-ctdbcpiolzma2x86_64-suse-linux~ₗE:pv ?N$?] b2sAXӦ{-W*b͓IPʑIχퟠHd)Nj^]bu

^/.翇\D:K@ǶIwkDŽ>ZW9]~A ufGOew(zC .Dxjx<q%džޤ@,JZ?vTV,JUmIi~?_*76I7e}>ͣQدg9B94t+rXa:`I1r@|̆:Y* fq/-806)8O%\-nOVfY x&pkbY̔0=(\{?8}1'TύO.h'c^?{N-'U_:Q42`7E>M->ևuB9 Bwź9kŁ0K9+F}/ m(t(#$ַv2y68Gs`׀,n`M(&Fŧ2+4l-֫]>Fei1V=gއ;v.2 vJJ凵6(Q=@ʃ*]d]{Rpel &|Y+`osDfm7JaP#wBI/cz~eĕB;yn302U?1=b4+Nq+~%EtF<ǩ=a!:O]4pSPFey#M RjCɩmR|evyibizEVet3bRSJp@Z.$k_ۺrv@ا׾dc0wQ, T!R#- q$>-Pʿ`'+geƻe8dLsDts9EN 67oI#U<-7l[$ۭ .=~vWҗ8czP*>ާ51f=隝/."?L7-nzX0,lBb 3olRvB K,c[~I#&GP)mrEUdG wL<2jϬKz5`$ |P _4fw'e0?1`жH}_KjhZt܇ XR5Y+ML;/C</hUaԚL5j-JSS߅!t ds؋S=`/6&[3yV, Qʿ64w6/#9&l.w5O&=_JHB9/=Hg5(aI&c{QҜUUsy!eg1FVWx6q ^уȍ_JbQ^a)&gԖV\k7RXh`qQyt8&HmF|Dj,^c6t.݉GPuY]A] ξATu )(#ԕ%YP=V/XjM/XL!uViim^ہ9Bo26&N~E6'ф,Di&!lNj xUSsO 8wjoܕ7~ט ;sfѤmgY6GG?n(F0<\X:#|f,^r-t2}&'6.Oq^sP'u_L4wYnِ+ %X0qƈK kץ9!:H[#ʹE↧!yVV ȄՌw@҈N='Jn[Ȏ+덜7eBu{jWn`m/WG+yC z4|MYacKw0P}ȼ _`oA{ 5Z2 AnMZE 8=)>s?;\X22d6=sO HG0ZU jۻg Io0ŗ%٘)%&C;nMՃ ፟&F{=Ku! @m"1"T}L zj6IZk/C5nvi)WS'wJPnIi&}5gesQv\;A 4 sqbꔵ6|Zq^=E >0ḩ5f39I:o0Y$_41?XT58p 0=#%{ )}JA^TcU>>']V~3kҤ]؇]_-F+%X* %S?=ʻ=UWz¥̎59{#y8!zt2nI Ik(t1zĈ9`9i_a|[VtOeg ~ ҤI|7OoW ֍~q@&ӭH #59 H;F& P:DzJ!UQJӯ]/[Tf@%$qѨeg$6cbwmD-p?- Uk̮',&S[(Kn+=>W?۩a~tJOU*NygӧE`^?rOp ("K*^$WZb}'M_˲Pˈ6uQڞP#馗KX9[_}4ikI$Q,M^6s|~i&CZtI])>p*@Sx"ac&`2+tLBͮC *6l~W]Llsvl9)˶OY0fp[?fS}u4Ud,Z|cG%7^29)CNɾEvřkWDǂ@rp!FoU^SX}wV*!xJ̤HcXkwr0٭N:heioQc5GzZ M͝_3KV!ĩ;<ݽ :UßTC/_2R ĵXeU+&KY /i>[yAu,Y]A̛t#d@D'œ$%"G6PqCՎW2J`7q!RÎS蓙2ƛV"ðnV%iz΢)mz:u1lTB;KF|[ʽ(ksjXIܦuzX:/"K^Rqo#kh3)Z+t3wCmЃ֑0wUf,+hkK.߮ZeD!nBIArNɛgJAv5Y, ͎%%B+UuK=;gOSì㭥 ELl\0azחXo?÷f):/ Vi ?gX /mtN{{6xV,~U6p ] @R)=w _wq)/%gD?';P*L P!SE\`gh4pU}:".kI1t] :ÁDUm L8a4wymY@iQ~0r?rۛ E>ndisEv]qnn6>z%wq&qHTGW8` ߸~ |Ե̆%\Rw\eI,d~JgGAR@EРN&Jq&@3DZkwpOλk>EY3NIJ+鬞V`GvzEW'[15C'\qrK7K~B&\9loaqu;[nw6Cs'dR<=gy+UјIe %~%"Yjv2@6O^T抛jW>#$>o/8*` ٢A $ŞQcah4Qi5SE3?P6 2B0+џX\d;Q.w$S5yPǩs+KIaJ72+*>9 67't@ῙH >϶R2W–B;bNN6[AWHvLM0iO˦++bPŨ;F`2$zo+68m³Po4Ax!h*|>y,-N(緩hE+_Mhur6;/TBh>wL+a 12+ġs=G5Y O䱲jp= -ؓ)q75`#y lpe/gpc~y6!WMŹz18-~8SF,59j )YpԹT}=c1V_CKQ͚bZ֕#XIn;6|8|AR=qAbKEC-\} f"lpKѴr,D>UЋ3N2k猠hnXq \3)͙^ -vHzK NX}3deeŏ0)m7N>qh QR VM]Zv|ԊOeuU`hZD9i&Rr [J=B!"̱5aNu궦$Zwr^!w+)PlMzi$^yo4T ALx50(XՀ~:SLR9{\oBj`&9uR[k,!]bg!PlEk64+Xhx?Nl;qVNiezJn$VPܧzDJ~ϯcҌ[ a| ľG[;k]K%]|š1Yz\QA{4˯-]Z k@+̿|Î~Vŕ?{=]:|ಱy- C!|+L_\UZ:![h7.`vb4Nt{_f~in@LĄ]?Xn!;äF cwĦr@a}-zϭ鬪Y>~1‰)Kj韅B-SmѴ6P. D67_% <=a ֜TxrCg26?(InC?RQR{/M_|9n]t@UW s]{O[3_(d Ձ염B#v>;խX=CDpD}EWXwn6k9 &LXV MbImuly1 lIǤl=TÊ(ʮ0&1I)5̓CGG*]{馮`40:Q1teZ6CJG=V ga D[E<]]c\?8P%zDębڈqjI"ag9H :NEM.Kז+ĿQThԒR ?D1)߼#3<=|dx@zRFèoCEt-c+Tuf? ~8 c@&DqE"ъv`nԚ9ܧ$o  [Y+1o@\(9"OFT]4N1sKO5E+@H9>|]#`z.(̷tN[.YYo"eXi ?ftȦV>@y`+xHۉۡEtRhavPwN#EQrfO {z7z[TB&1٬ZܡS ܼ Ȩ^L2PDA $~T;qCbCXC݄YBZ>$_swud !煉Xf#6цsblw$[8Xba{+(fG3&/Fe3&O`bB`<9*Qy.a_ Ov-`LC|7FuK.4⨧i6z & GqCS&& Uh,޺#bC ̎.y@@{#zq(<ˆMHQ)gMȊG礩}] 4׆Ǝ*HĴoO\2=d U:8_2YP}jn,Jz$ q =(e˅>;yI#PRV@ʤA0af!Uk lfB4#>BchoIm9&jo:ߗ'ܩ4O͵6CڰA2V#h%@K2x0%p.X/*/K <6M^%oņ\nSɧD0N%CSu;ˇkǁ$Q~N-/JOE QA>ltYu9*e=i4&vٰV2M."v@Q.玐Y;e7zf I E\H*GBud`J@#/HyΥ^᫢إ ѲY`4ӽiMI.Kg FAQt|+59t+S j i[z[ꆯʷM+;O?N%`f)AcS(At ۤÈ GQ`giW W&lDM_qAU՚WOlړG'҄g1sH_Do5ҀL~.{:J{^@%fԚIS~Aui\+|eXN&&̑E}h=|c{ձC"x2!W*V#H~~K̎x/{^VYd R7˜⋃$Irhpr2fMH6I tf&K.r#>jA k7Ƨ(vɷy9mc]qGg$JBkl5 h!Knm7c'JAh2,vV{fH@w8醮_l40담c:NEMߍ," {†3i0Ng,{xp]3m#KKeÍ@ťJߗb\95tgebcjSҦ&kt>gGУ_+i4R_!uglBXW/mЏrfbH\q& v#9U>BX8K-l)Y:Rxv[\F$`dJh* 齾3Usf~B8mbGsǠzN'S/uF\Kt!C`d-o5Jv#d-5,AM ]ZFvyt>}+pyH*M|6-s?l2n_Me8cf0"]J#u}ɕбfZ{,jb-e{m#Qc,phcDZ?ét~ KQIHJ-s>^sKح388I&^_uo!vzv׋ SQ8:8C(+})Yw3(A눶;[#keÇ"fV}gM.\ 8׳WďmhqwF>4Fm$۝q~$J ƶ梉q?u'mC̼ңߊuio F LߨmLk8|!֊3F]u9PKXk8PBЉreQ[2x"񠀆w"HJ31oմ|p˫}ʵy-Ҳf 8C2l&ݦ$.{z-IG[ZdQ۫ rāJb,kRi@^BY:+h|g8`&„$Y!8e~FC/ Ɍ~MHAUbU_& B>:,y66(}c ; B( TW QhɏX~X;lm;+ٖ+eY(۔@%6E@XmC5Fmܲ\K%#NIƝEg ' vwiW0Z;< T>:q nYU<XNPdCAһe&cU8W!{y^.'AK__ 2eV'bRHEvPjvm̂aflIsg z{_ a7O!5D cu9B%3wxtM9_.FSς*tPKgVZiegz+:leJMgGd_ULP`BܼOJ>Pր6]mD-:>l<{R6 |E{+{? \"g* q CVp1 pJXMW'lQD-dVhs®.t">8 ͮ3D^mgR(W6[%IL ju{e2Bbݡu^4"mYVl ɨ`fHARb=$H1dL.m/ y~DdpgKf)f%2F6vxTHFYB=_!(JF;|u(Ew;^.kP]DRfmjQ ^V?#F xxmĒ=8^ݯKҢ-o"SSdK9mٝːh hh/Mw0Q˘a7T jK~Ͽ15?<۽!4YC//GF17YBH{1 '׎/;]zO/kf9՘x3 nDL ^=볭 M6puuIϔI꼃wU2PF.NjuG^R䛠%rDz>cG᧝Bt|G>T@n,F1}bE"!= W7ўiRP[5C̾O."TNBN*BlR\Hdc$ 9p r`(*9M"0p[f%$i[υ+5gmVD-~¯0 %@HD6pxX%e7+/K]4 \7iX&Oo#.VX5>(#,paEͰ:w)CioT^-HؒRr(tf[{eAbh܋vĪqRv ){Zțd]ǷT˖]jW.'.EH`K". [=ɜ-\du SP}-BGV0ob9~ 'o1VSC2eg+ m2o )}a6=TI$.ʥB'?f$`q@Q['{%Jk] _Fm~<٧W'ip?š CNLAbYM.'j ԬE˗[62Qo_;m/nTJLDCڱfR{W1}{d۪[eX'}ՋjA/Y t Z3M5܁BPèGuT2s$OݣˤX9E%&lC:B4::z30fװ1h_HhKVSR *ESqY8r:F3L\ˣIXw'87:RL~ bHQ ) 'Bgb";lj; 񇢘*.P3{O`޳QltՌR)?e_2Ev=SƓ)b"*7W<9XJI$ƼUdxmuWMRJNN d`s~l Ѫ =dd%u7Z},!:Ч /gAV c! hi JSeE{RX5b~цDiSC_B nƈ{+i)M -zrlbH(Fd`Wq&'w}| Vܥ4Ä>n-U8er,:dp= H:H4(whE3j('de|y%0b=Wa8PYj>EokHN)Жٮoa<1# Wa-񷚠TRp]YEu_}w mG9rcs3LQ45nnclT%?fpw W<[|DU0!+bZ/M^(S-wxR_ t|F4Laýz7#ۅťI'ss'