Search Results

Search found 1815 results on 73 pages for 'percona xtradb cluster'.

Page 62/73 | < Previous Page | 58 59 60 61 62 63 64 65 66 67 68 69  | Next Page >

  • To clone or to automate a system installation?

    - by Shtééf
    Let's say you're setting up a cluster of servers performing the same task. Or say you're just setting up a bunch of different servers, but you expect to use a base configuration on all of your servers. Would it be better practice to create a base image and clone it, or to automate the installation and configuration? I occasionally end up in this argument with my boss, in situations where we're time-pressed. When he sees me struggle with perfecting the automation, his suggestion is often to clone the entire disk to the other machines. But my instinct has always been to avoid cloning. This is mostly from an Ubuntu perspective, but the question is fairly general. My reasons for avoiding cloning are: On a typical install, even if it's fresh, there are already several unique identifiers installed: filesystem UUIDs, SSH host keys, among others. These would have to be regenerated. Network needs to be reconfigured for each clone. This would need to be done off-line, of course, or the settings will conflict with other machines on the network. On the other hand, some of the cloning advantages are quite clear as well: (Initially?) less effort required than automating configuration. Tools exist to quickly address (some) of the above disadvantages. (I can see right through my own bias there.)

    Read the article

  • Cassandra Remote Connection

    - by Lyuben Todorov
    I'm not managing to connect to cassandra from outside machines. The database is hosted on a windows machine and im trying to connect through a mac (but this shouldn't cause problems) Local connection works: C:\cassandra\bin>cassandra-cli Starting Cassandra Client Connected to: "Test Cluster" on 127.0.0.1/9160 Welcome to Cassandra CLI version 1.1.6 But fails from other machines on the same network bin/cassandra-cli --host 192.168.0.10 --port 9160 org.apache.thrift.transport.TTransportException: java.net.ConnectException: Operation timed out at org.apache.thrift.transport.TSocket.open(TSocket.java:183) at org.apache.thrift.transport.TFramedTransport.open(TFramedTransport.java:81) at org.apache.cassandra.cli.CliMain.connect(CliMain.java:70) at org.apache.cassandra.cli.CliMain.main(CliMain.java:246) Exception connecting to 192.168.0.10/9160. Reason: Operation timed out. Welcome to Cassandra CLI version 1.2.0-beta3 Type 'help;' or '?' for help. Type 'quit;' or 'exit;' to quit. There is a router on the network but these ports have been triggred Ports: 1024, 7000, 7001, 7199, 9160 And the same ports were forwarded to 192.168.0.10 (where Cassandra is hosted) Cassandra version is 1.0.7 And the settings I think i need to change in cassandra.yaml listen_address: 192.168.0.10 rpc_address: I'm not really sure if I've missed any steps. Any help would be appreciated.

    Read the article

  • High mysql server load, sar output

    - by eric
    I have a MySQL Server that should be performing better than it seems to be. We're running ubuntu on a Amazon Cluster Compute (cc1.4xlarge) Linux ip-10-0-1-60 3.2.0-25-virtual #40-Ubuntu SMP Wed May 23 22:20:17 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux Distributor ID: Ubuntu Description: Ubuntu 12.04 LTS Release: 12.04 Codename: precise I have several output files from sar that i'm not really sure how to interpret. For example, I ran: # Individual block device I/O activities sar -d 1 180 > logs/block_device_io.log & which gave me what looks like really high utilisation of my disk (turns out this block device maps to /dev/xvdh on /var/lib/mysql type ext4 (rw,_netdev) The output from my log: 10:48:59 PM DEV tps rd_sec/s wr_sec/s avgrq-sz avgqu-sz await svctm %util 10:49:00 PM dev202-16 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 10:49:00 PM dev202-32 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 10:49:00 PM dev8-0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 10:49:00 PM dev202-112 1008.00 31040.00 1416.00 32.20 1.02 1.01 0.89 90.00 10:49:00 PM dev202-80 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Am I wrong in thinking this is a problem? I have it above 90% almost the entire time we're seeing slowness. Or does this just mean MySQL is doing what it's supposed to do?

    Read the article

  • Nexus 1000v VEM fails on 2 out of 8 hosts.

    - by cougar694u
    I have 8 ESXi hosts. I do a fresh install from the installable CD directly to 4u1. We have another 2-node cluster with a working Nexus 1000v primary & secondary. Everything's up and running. I installed 6 hosts and everything worked great, migrated them to the Nexus DVS, and VUM installed the modules. I did the 7th host, and when I tried to migrate it to the DVS, it failed with the following error: Cannot complete a Distributed Virtual Switch operation for one or more host memebers. DVS Operation failed on host , error durring the configuration of the host: create dvswitch failed with the following error message: SysinfoException: Node (VSI_NODE_net_create) ; Status(bad0003)= Not found ; Message = Instance(0): Inpute(3) DvsPortset-0 256 cisco_nexus_1000v got (vim.fault.PlatformConfigFault) exception Then, I tried to do host 8, and got the exact same problem. It worked about 15 minutes prior when I did host 6, nothing changed, then went to host 7 and it failed. If I try to remediate either of these two hosts, either patches or extensions, it fails. Anyone else have these problems?

    Read the article

  • Backing up SQL NetApp Snapshots using TSM

    - by WerkkreW
    In our environment we have a 3 node SQL 2005 Cluster which is on NetApp storage. We are currently using SMSQL (NetApp SnapManager for SQL) to take Snapshot backups of the data. This works great, but due to some audit requirements we are also forced to maintain some copies on tape. We have used NDMP in other places across the enterprise but we do not want to use it in this specific instance. Basically what I need to do is, get the most recent snapshot copy of the databases on tape, via Tivoli Storage Manager (TSM). What I have done is, obtained a basic Windows Server 2003 VM with SnapDrive installed, which is SAN attached and zoned to the NetApp, and I have written a batch file to do the following: Mount the latest __RECENT snapshot lun to the host, using a specific drive letter Perform a TSM based incremental backup Dis-mount the LUN This seems to work fine, except sometimes the LUN's do not mount due to some sort of timeout. Also, due to my limited knowledge of windows batch scripting, I have no way to monitor the success or failure of these backups since I do not know how to send a valid return code back to the TSM scheduling service. Is there a more efficient/elegant way to accomplish this without NDMP?

    Read the article

  • ScreenOS MIP selection for outbound connections

    - by David Mackintosh
    Given a ScreenOS 6.3.0 firewall with this configuration: unset flow reverse-route clear-text set interface "ethernet0/0" zone "Trust" set interface ethernet0/0 ip 192.168.1.1/24 set interface ethernet0/2 ip 10.0.0.1/24 set interface ethernet0/2 mip 10.9.9.10 host 192.168.1.10 netmask 255.255.255.255 vr trust-vr set interface ethernet0/2 mip 10.8.8.10 host 192.168.1.10 netmask 255.255.255.255 vr trust-vr set route 0.0.0.0/0 interface ethernet0/2 gateway 10.0.0.254 ... (and assuming appropriate policies) is there a way to control which MIP the firewall selects for outbound connections initiated by 192.168.1.10? (presumably because I want the upstream 10.0.0.254 to do different things with the packet depending on what its source IP is). Update: OK, my scenario is this: I have two ISP uplinks, each giving me a different IP space. Each ISP will only route the IP space they have assigned me. So in order for the server to be reachable by both ISPs, I need two MIPs for the same server, one on each interface. Inbound traffic will work fine with the 'unset flow reverse-route clear-text' option in use. But for outbound (think an email server sending a message) I need a way to select an appropriate MIP depending on which outbound link I want the system to use, and I want this to be resiliant so that if the link the firewall prefers goes down overnight I don't have to manually flip something to keep the mail flowing. Basically I'm trying to avoid having to buy a link-balancer appliance (or two since I'll need a cluster). Is there a way to do this?

    Read the article

  • Wiping Deleted Directory Entries and Defragmenting Directories

    - by Synetech inc.
    Hi, I have seen plenty of apps that wipe free space on a disk (usually by creating a file that is as big as the remaining space) or defragment a file (usually by using the MoveFile API to copy it to a new contiguous area). What I have not seen however is a program that wipes the deleted directory entries. That is, when a file is deleted, its information (name, dates, etc.) remain in the directory, but are simply marked as empty. That leaves all kinds of information in a directory entry, and also wastes space since (at least on FAT drives), the directory may be using several clusters. For example, if a directory once had a lot of files, it will be expanded to use another cluster which could be anywhere on the disk. This means that the directory is fragmented, and may be using more clusters than needed, possibly with 100’s of unused (ie, “deleted file”) entries between active files. Does anyone know of a program that can defragment/consolidate directories (ie, wipe unused entries, and move active entries together)? (I would really rather not have to resort to writing my own yet again.) Thanks a lot. EDIT Sorry, I should have said, Windows and/or DOS, for FAT*/NTFS.

    Read the article

  • How do I make dnsmasq serve IP addresses via IPoIB?

    - by Matt
    I have a cluster farm that I'm setting up. The nodes (computers in the farm) are connected via ethernet & IP over Infiniband. I'm needing to netboot the nodes and thought dnsmasq would fit well as it provides all the features including support for DHCP over IB and it works great for our ethernet setup. However, I can't seem to get it to provide IP addresses to the infiniband adaptors on the nodes. Each node is running an Ubuntu desktop 12.04 LTS. The dnsmasq server is running on ubuntu server 12.04LTS and has the following test config: dhcp-authoritative domain-needed bogus-priv expand-hosts no-hosts domain=local dhcp-range=eth0,10.0.0.10,10.0.0.255,12h dhcp-option=eth0,3,10.0.0.1 dhcp-range=ib0,10.1.1.10,10.1.1.255,12h dhcp-option=ib0,3,10.1.1.1 log-queries log-dhcp IPoIB works between nodes when configured statically but not with dhcp. On the nodes the file /etc/network/interfaces contains auto lo iface lo inet loopback auto ib0 iface ib0 inet dhcp #iface ib0 inet static #address 10.1.1.5 #netmask 255.0.0.0 up echo connected >`find /sys -name mode | grep ib0` Is there something I need to do on the client or server end to make this work?

    Read the article

  • vmware vmdk disk problem

    - by dmtr
    I have a VMware ESXi 4 server and 2 storage servers (mounted via nfs). Between the storage servers (Fedora 14) is a drbd cluster (dual primary) and ocfs2 filesystem; also every server has a local partition with an ext4 filesystem, both are mounted via nfs on the esxi server. When I tried to copy a virtual machine (naturally it was powered off) from the ext4 partition to the ocfs2 partition, the vmdk total file size is different, but the md5sum is the same. On the ext4 partition: # ls -la total 28492228 -rw------- 1 root root 42949672960 Jan 14 14:46 disk-flat.vmdk # md5sum disk-flat.vmdk 0eaebe3138beb32f54ea5de6dfe5a987 On the ocfs2 partition: # ls -la total 13974660 -rw------- 1 root root 42949672960 Jan 14 16:16 disk-flat.vmdk # md5sum disk-flat.vmdk 0eaebe3138beb32f54ea5de6dfe5a987 When I power on the virtual machine from the ocfs2 partition it dosn't work. I have a windows on the virtual machine and it freez?s after the windows logo. From the ext4 partition the virtual machine workes. I tested with linux (created and installed on ext4 partition and then copied to the ocfs2) and the same problem appears. When I create a virtual machine directly from ocfs2 partition, there are no problems. I tried to copy via vSphere client, and I have the same problem. Any suggestions?

    Read the article

  • How to deduplicate 40TB of data?

    - by Michael Stauffer
    I've inherited a research cluster with ~40TB of data across three filesystems. The data stretches back almost 15 years, and there are most likely a good amount of duplicates as researchers copy each others data for different reasons and then just hang on to the copies. I know about de-duping tools like fdupes and rmlint. I'm trying to find one that will work on such a large dataset. I don't care if it takes weeks (or maybe even months) to crawl all the data - I'll probably throttle it anyway to go easy on the filesystems. But I need to find a tool that's either somehow super efficient with RAM, or can store all the intermediary data it needs in files rather than RAM. I'm assuming that my RAM (64GB) will be exhausted if I crawl through all this data as one set. I'm experimenting with fdupes now on a 900GB tree. It's 25% of the way through and RAM usage has been slowly creeping up the whole time, now it's at 700MB. Or, is there a way to direct a process to use disk-mapped RAM so there's much more available and it doesn't use system RAM? I'm running CentOS 6.

    Read the article

  • VMware vSphere 4.1 and BackupExec 2010

    - by Josh
    I'm sure a common problem with most shops is backups, their size, and the window in which you have to back up the data. What we are working with: VMware vSphere 4.1 Cluster PS4000XV Equallogic Storage Array (1.6TB Volume dedicated for Backup to Disk) Physical Backup Server with a single LTO4 drive. BackupExec 2010 R3 with the following agents, Exchange, SQL, Active Directory, VMware. Dual Gigabit MPIO Connections between all devices (Storage Array, Backup Server, VM Hosts) What we would like to accomplish: I would like to implement an efficient Backup to Disk to Tape solution where all of our VMs are backed up to the Storage Array first, and then once completely backed up to the array are replicated to tape. In the event we needed to recover, we would be able to do so directly from tape. Where we are at currently. Of the several ways I have setup the jobs in Backup Exec 2010 R3 the backup jobs all queue up at the same time, as soon as a job is finished backing up to disk it then starts that same job to tape, but pulling from the original source instead of the designated B2D location. I understand that I could create a job that backs up the "Backup to Disk" folder to tape, but in the event of restoration, I would first need to stage the data in the B2D folder before I could restore the VM. I would really like to hear from individuals in similar situations. Any and all comments and critiques are appreciated.

    Read the article

  • SQL 2008 Replication corrupt data problem

    - by Jonathan K
    We took a SQL 2000 database. Took a lightspeed backup. Restored on SQL 2008 active/passive cluster. Then setup replication to replicate the data back to SQL 2000. So 2008 is the publisher/distributor, and 2000 is doing a pull subscription. Everything works well, execpt we occassionally get corrupt data in varchar/text fields on the subscriber. So for example we have a table with 4500 records. When we run this statement: update MedstaffProvider set Notes = 'Cell Phone: 360.123.4567 Answering Service: 360.123.9876' where LastName = 'smith' The record in the 2008 database is updated as expected. But in the subsriber datbase we'll get gibberish in the notes field: óPÌ[1] T $Oé[1] ð²ñ. K Here's what we know: This is repeatable, meaning we can run that same query all day long and get the same gibberish. If you alter update statement slightly the data gets replicated just fine. The collation on both databases is the same. So far we've only detected the problem with text/varchar fields. (The notes field above is text). Only one or two records in a table are impacted. The table structure looks identical in both 2000/2008. We haven't made any changes. We have found one solution that fixes the problem. Basically if we recreate the table in 2008 (say as MedStaffProvider2) and then insert all the data. Drop the original table. Rename the table to it's original name. Setup replication again. And run the exact same update statement it works as expected. Does anyone have any idea what might be happening here? Or are there any other techniques we can use to troubleshoot this? I've found a solution for this, but would really like to undertsand why this is happening.

    Read the article

  • Apache process consuming all memory on the server

    - by jemmille
    I have an apache process that suddenly appears on a particular server. When it shows up it starts consuming memory at a very rapid rate, then moves on to all the swap. In all it consumes about 11GB (including swap) of memory and the server eventually becomes unresponsive. The load on the server is under 1 at all other times. The process runs as nobody and I am having a hard time tracking down the source. If i run an strace on the process and all it did was continuously dump out mprotect over and over again If i run lsof -p <pid>, I get this, but only sometimes: httpd 19229 nobody 152u IPv4 175050 crawl-66-249-67-216.googlebot.com:62336 (CLOSE_WAIT) httpd 19229 nobody 153u IPv4 179104 crawl-66-249-71-167.googlebot.com:58012 (ESTABLISHED) As long as I catch it, I can kill the process and the server almost immediately stabilizes. I have on site on the server that is getting a few thousand hits a a day that I think might be the source, but I still can't find the exact reason. Also, this is a cPanel server and I have upcp'd the server, rebuilt apache with easy apache, and rebuilt httpd.conf. It is not spawing any related processes, meaning I can find any php, mysql, cgi, etc. processes that relate to this process. It's just a loner process that balloons fast and consumes ever last MB of memory. This is on a XenServer 5.6 based VM. No other servers in the cluster are having this issue.

    Read the article

  • Best practices for settings for Oracle database creation

    - by Gary
    When installing an Oracle Database, what non-default settings would you normally apply (or consider applying) ? I'm not after hardware dependent setting (eg memory allocation) or file locations, but more general items. Similarly anything that is a particular requirement for a specific application rather than generally applicable isn't really useful. Do you separate out code/API schemas (PL/SQL owners) from data schemes (table owners) ? Do you use default or non-default roles, and if the latter, do you password protect the role ? I'm also interested in whether there's any places where you do a REVOKE of a GRANT that is installed by default. That may be version dependent as 11g seems more locked down for its default install. These are ones I used in a recent setup. I'd like to know whether I missed anything or where you disagree (and why). Database Parameters Auditing (AUDIT_TRAIL to DB and AUDIT_SYS_OPERATIONS to YES) DB_BLOCK_CHECKSUM and DB_BLOCK_CHECKING (both to FULL) GLOBAL_NAMES to true OPEN_LINKS to 0 (did not expect them to be used in this environment) Character set - AL32UTF8 Profiles I created an amended password verify function that used the apex dictionary table (FLOWS_030000.wwv_flow_dictionary$) as an extra check to prevent simple passwords. Developer logins CREATE PROFILE profile_dev LIMIT FAILED_LOGIN_ATTEMPTS 8 PASSWORD_LIFE_TIME 32 PASSWORD_REUSE_TIME 366 PASSWORD_REUSE_MAX 12 PASSWORD_LOCK_TIME 6 PASSWORD_GRACE_TIME 8 PASSWORD_VERIFY_FUNCTION verify_function_11g SESSIONS_PER_USER unlimited CPU_PER_SESSION unlimited CPU_PER_CALL unlimited PRIVATE_SGA unlimited CONNECT_TIME 1080 IDLE_TIME 180 LOGICAL_READS_PER_SESSION unlimited LOGICAL_READS_PER_CALL unlimited; Application login CREATE PROFILE profile_app LIMIT FAILED_LOGIN_ATTEMPTS 3 PASSWORD_LIFE_TIME 999 PASSWORD_REUSE_TIME 999 PASSWORD_REUSE_MAX 1 PASSWORD_LOCK_TIME 999 PASSWORD_GRACE_TIME 999 PASSWORD_VERIFY_FUNCTION verify_function_11g SESSIONS_PER_USER unlimited CPU_PER_SESSION unlimited CPU_PER_CALL unlimited PRIVATE_SGA unlimited CONNECT_TIME unlimited IDLE_TIME unlimited LOGICAL_READS_PER_SESSION unlimited LOGICAL_READS_PER_CALL unlimited; Privileges for a standard schema owner account CREATE CLUSTER CREATE TYPE CREATE TABLE CREATE VIEW CREATE PROCEDURE CREATE JOB CREATE MATERIALIZED VIEW CREATE SEQUENCE CREATE SYNONYM CREATE TRIGGER

    Read the article

  • Advice needed: warm backup solution for SQL Server 2008 Express?

    - by Mikey Cee
    What are my options for achieving a warm backup server for a SQL Server Express instance running a single database? Sitting beside my production SQL Server 2008 Express box I have a second physical box currently doing nothing. I want to use this second box as a warm backup server by somehow replicating my production database in near real time (a little bit of data loss is acceptable). The database is very small and resources are utilized very lightly. In the case that the production server dies, I would manually reconfigure my application to point to the backup server instead. Although Express doesn't support log shipping natively, I am thinking that I could manually script a poor man's version of it, where I use batch files to take the logs and copy them across the network and apply them to the second server at 5 minute intervals. Does anyone have any advice on whether this is technically achievable, or if there is a better way to do what I am trying to do? Note that I want to avoid having to pay for the full version of SQL Server and configure mirroring as I think it is an overkill for this application. I understand that other DB platforms may present suitable options (eg. a MySQL Cluster), but for the purposes of this discussion, let's assume we have to stick to SQL Server.

    Read the article

  • How to migrate a running KVM (with full disk copy) to another node?

    - by klipz
    I'm doing tests on KVM, and I'd like to see if I can make a hot migration, I mean the virtual machine won't stop running during the migration (but a few seconds of freeze is ok). I use a small cluster for my test : kvm1, kvm2, and kvmnfs. kvm1 and kvm2 runs the virtual machines kvmnfs is a NFS server, and it's mounted on /KVM on both kvm1 and kvm2 To migrate a VM (only RAM in fact) from kvm1 to kvm2, I run the same kvm command on kvm2 (with -incoming tcp:0:4444) that on kvm1, then I use "migrate -d tcp:kvm2:4444" : It works great, since the VM file is common to both machines. Now, I wan't to make a full migration (RAM + disk) of a local VM file (no more NFS) of kvm1 to kvm2. I tried to create an empty file, with touch, on kvm2 and use the same kvm command line + the "-incoming ..."). Then on kvm1 I use "migrate -d tcp:kvm2:4444" : It copies everything, then... the VM fails (any I/O disk gives an I/O error) ! And my VM file on kvm2, the one I created with touch, as still a size of 0 bytes. What am I doing wrong ? What is the exact command to use on kvm2 ? And what is the command to launch, in the monitoring mode, on kvm1 ?

    Read the article

  • Finding cause of TCP retransmission within a LAN

    - by Surreal
    Hello denizens of Server Fault I have an irritating problem with a LAN of about 100 computers, 2 Windows domain servers, and 12 VoIP phones. Since their installation around a year ago, every week or so, we notice a VoIP phone resetting itself - occasionally in the middle of a call. Simultaneously there are often signs of temporary loss of connection on computers: freezes in explorer while accessing network shares, errors in our administration software due to loss of connection to the database server. I have been doing some Wireshark monitoring on the connection between the VoIP PBX and the rest of the network. Wireshark picks up a clump of retransmitted TCP packets at the times when we record phone restarts. The Wireshark log shows about 2 clusters of retransmissions a day ranging from 5 packets to hundreds. Those in each cluster are mainly between the PBX and some set of the VoIP phones, but not always the same set. Often retransmissions at the same time are to phones connected to the same switch, but sometimes retransmissions occur together to phones at opposite ends of the network. There are usually some coincident retransmissions in passing TCP traffic, for example between client machines and the file servers. The spikes in retransmissions and phone resets do not correlate well with when the network is heavily loaded. They seem to occur slightly more during the day, but most in the evening, when traffic should be decreasing. They occur reasonably often late at night when most computers are turned off and traffic should be lowest. Do you have any ideas that might help diagnose the cause of problems like this? One thing I have not yet tried, but should have, is updating the firmware of all the switches.

    Read the article

  • Problem with diacritics on psql 9.0 (PostgreSQL)

    - by Gaks
    I have two instances of PostgreSQL installed on my server: 8.3 and 9.0. There seams to be some problem with Polish diacritic characters (like óleaszzc) on postgresql 9.0 client - psql. When I connect to DB (either 8.3 or 9.0) with psql 8.3 - I can type all diacritics on the terminal without any problems: www:/tmp# sudo -u postgres /usr/lib/postgresql/8.3/bin/psql -q postgres=# ólscn However, when I connect to the same DBs with psql 9.0 client - I can't type diacritics on the terminal anymore: www:/tmp# sudo -u postgres /usr/lib/postgresql/8.3/bin/psql -q Here are some encoding settings: www:/tmp# sudo -u postgres /usr/lib/postgresql/9.0/bin/psql -q -c "show client_encoding" client_encoding ----------------- UTF8 (1 row) . www:/tmp# sudo -u postgres /usr/lib/postgresql/8.3/bin/psql -q -c "show client_encoding" client_encoding ----------------- UTF8 (1 row) . www:/tmp# sudo -u postgres /usr/lib/postgresql/9.0/bin/psql -q -l List of databases Name | Owner | Encoding | Collation | Ctype | Access privileges ---------------------+--------------+----------+-------------+-------------+----------------------- postgres | postgres | UTF8 | pl_PL.UTF-8 | pl_PL.UTF-8 | . www:/tmp# echo $LANG pl_PL.UTF-8 It looks like DB/cluster configuration doesn't matter - if psql 8.x on terminal works fine and psql 9.x does not. Any idea how to fix that?

    Read the article

  • Postfix Whitelist before recipient restrictions

    - by GruffTech
    Alright. Some background. We have an anti-spam cluster trucking about 2-3 million emails per day, blocking somewhere in the range of 99% of spam email from our end users. The underlying SMTP server is Postfix 2.2.10. The "Frontline defense" before mail gets carted off to SpamAssassin/ClamAV/ ect ect, is attached below. ...basic config.... smtpd_recipient_restrictions = reject_unauth_destination, reject_rbl_client b.barracudacentral.org, reject_rbl_client cbl.abuseat.org, reject_rbl_client bl.mailspike.net, check_policy_service unix:postgrey/socket ...more basic config.... As you can see, standard RBL services from various companies, as well as a Postgrey service. The problem is, I have one client (out of thousands) who is very upset that we blocked an important email of theirs. It was sent through a russian freemailer who was currently blocked in two of our three RBL servers. I explained the situation to them, however they are insisting we do not block any of their emails. So i need a method of whitelisting ANY email that comes to domain.com, however i need it to take place before any of the recipient restrictions, they want no RBL or postgrey blocking at all. I've done a bit of research myself, http://www.howtoforge.com/how-to-whitelist-hosts-ip-addresses-in-postfix seemed to be a good guide at first, almost fixing my problem, But i want it to accept based on TO address, not originating server.

    Read the article

  • Router failover not detecting outside interface link lost

    - by Matt
    Suppose I have two routers configured in master/slave configuration. They look something like this (addresses are not real ones) 123.123.123.10 <===> [eth0] Router 1 (10.1.1.2) [eth1] ===> +----------+ | 10.1.1.1 | ===> LAN 172.123.123.10 <===> [eth0] Router 2 (10.1.1.3) [eth1] ===> +----------+ The 10.1.1.1 is the default route for the Network (10.1.1.0). What's slightly different in this config to other's I've seen is that I don't have an external virtual IP. Also, the 10.1.1.1 addresses are in real life, public IP's (not private ones shown here). This is more of a router setup than a firewall setup so I'm not using NAT here. Now the issue that I'm having is that I can't see any way to configure UCARP or VRRP to monitor both eth0 & eth1 and fail over to the backup router should either of them go down. What I'm seeing is that if Router1 is the master and I unplug eth0 on router1, it doesn't fail over to router 2. However, it will if instead I unplug eth1 of router 1. In VRRP I see there is a cluster group, but it seems that for this to work you need to have virtual ip's or vrrp instances rather than actual interfaces assigned to it. I hope my explanation is clear. How do I get around this?

    Read the article

  • RabbitMQ message consumers stop consuming messages

    - by Bruno Thomas
    Hi server fault, Our team is in a spike sprint to choose between ActiveMQ or RabbitMQ. We made 2 little producer/consumer spikes sending an object message with an array of 16 strings, a timestamp, and 2 integers. The spikes are ok on our devs machines (messages are well consumed). Then came the benchs. We first noticed that somtimes, on our machines, when we were sending a lot of messages the consumer was sometimes hanging. It was there, but the messsages were accumulating in the queue. When we went on the bench plateform : cluster of 2 rabbitmq machines 4 cores/3.2Ghz, 4Gb RAM, load balanced by a VIP one to 6 consumers running on the rabbitmq machines, saving the messages in a mysql DB (same type of machine for the DB) 12 producers running on 12 AS machines (tomcat), attacked with jmeter running on another machine. The load is about 600 to 700 http request per second, on the servlets that produces the same load of RabbitMQ messages. We noticed that sometimes, consumers hang (well, they are not blocked, but they dont consume messages anymore). We can see that because each consumer save around 100 msg/sec in database, so when one is stopping consumming, the overall messages saved per seconds in DB fall down with the same ratio (if let say 3 consumers stop, we fall around 600 msg/sec to 300 msg/sec). During that time, the producers are ok, and still produce at the jmeter rate (around 600 msg/sec). The messages are in the queues and taken by the consumers still "alive". We load all the servlets with the producers first, then launch all the consumers one by one, checking if the connexions are ok, then run jmeter. We are sending messages to one direct exchange. All consumers are listening to one persistent queue bounded to the exchange. That point is major for our choice. Have you seen this with rabbitmq, do you have an idea of what is going on ? Thank you for your answers.

    Read the article

  • how to design pound -> varnish -> jboss for ha + loadbalancing

    - by andreash
    Hello, I'm planning a new infrastructure for our web application. We have two JBossAS5 servers, running in a cluster. Session state will be replicated via JBoss Cache. In front of that, there should be some cache, to speed up delivery of static elements. However, most of the traffic to our app will be via HTTPS. So far, I had been thinking of two Varnish caches in front of the JBossASs, each being configured for loadbalancing to the two JBossASs via round-robin. Since Varnish doesn't handle HTTPS, then there would need to be two pound proxies in front of the Varnishs, dealing with the HTTPS. The two pounds would be made high-available with Heartbeat/LinuxHA. The traffic to www.example.com would then be going through our firewall, from there to the virtual IP of the pounds, from there to the Varnishs, and from there to the JBossASs. Question 1: Does this make sense? Or is it overly complicated, and the same goal can be reached with simpler methods? Question 2: If my layout is fine, how do I configure the pound - Varnish step? Should I a) make the Varnish service high-available through Heartbeat/LinuxHA as well and direct traffic from pound to the virtual IP of the Varnishs, or should I rather b) Configure two independent Varnishs and use load-balancing in pound to address the different Varnishs? Thanks a lot for your insight! Andreas.

    Read the article

  • Problem posting multipart form data using Apache with mod_proxy to a mongrel instance

    - by Ryan E
    I am attempting to simulate my site's production environment as closely as I can on my local machine. This is a rails site that uses Apache w/ mod_proxy to forward requests to a mongrel cluster. On my Mac OSX Leopard machine, I have the default install of apache running and have configured a vhost to use mod_proxy to to forward requests to a local running mongrel instance on port 3000. <Proxy balancer://mongrel_cluster-development> BalancerMember http://127.0.0.1:3000 </Proxy> For the most part, this is working fine. I can browse my development site using the ServerName of the vhost I configured and can confirm that requests are being properly forwarded to the mongrel instance. However, there is a page on the site that has a multipart form that is used to upload an image to the server. When I post this form, there is a delay of about 5 minutes and the browser ultimately returns a Bad Request Your browser sent a request that this server could not understand. In the error log for my vhost: [Tue Sep 22 09:47:57 2009] [error] (70007)The timeout specified has expired: proxy: prefetch request body failed to 127.0.0.1:3000 (127.0.0.1) from ::1 () This same form works fine if I browse directly to the mongrel instance (http://127.0.0.1:3000). Anybody have any idea what the problem might be and how to fix it? If there is any important information that I neglected to include, post a comment, and I can add to this question. Note: Upon further investigation, this appears to be a problem specific to Safari. The form works fine in Firefox.

    Read the article

  • Architecture for highly available MySQL with automatic failover in physically diverse locations

    - by Warner
    I have been researching high availability (HA) solutions for MySQL between data centers. For servers located in the same physical environment, I have preferred dual master with heartbeat (floating VIP) using an active passive approach. The heartbeat is over both a serial connection as well as an ethernet connection. Ultimately, my goal is to maintain this same level of availability but between data centers. I want to dynamically failover between both data centers without manual intervention and still maintain data integrity. There would be BGP on top. Web clusters in both locations, which would have the potential to route to the databases between both sides. If the Internet connection went down on site 1, clients would route through site 2, to the Web cluster, and then to the database in site 1 if the link between both sites is still up. With this scenario, due to the lack of physical link (serial) there is a more likely chance of split brain. If the WAN went down between both sites, the VIP would end up on both sites, where a variety of unpleasant scenarios could introduce desync. Another potential issue I see is difficulty scaling this infrastructure to a third data center in the future. The network layer is not a focus. The architecture is flexible at this stage. Again, my focus is a solution for maintaining data integrity as well as automatic failover with the MySQL databases. I would likely design the rest around this. Can you recommend a proven solution for MySQL HA between two physically diverse sites? Thank you for taking the time to read this. I look forward to reading your recommendations.

    Read the article

  • Shared firewall or multiple client specific firewalls?

    - by Tauren
    I'm trying to determine if I can use a single firewall for my entire network, including customer servers, or if each customer should have their own firewall. I've found that many hosting companies require each client with a cluster of servers to have their own firewall. If you need a web node and a database node, you also have to get a firewall, and pay another monthly fee for it. I have colo space with several KVM virtualization servers hosting VPS services to many different customers. Each KVM host is running a software iptables firewall that only allows specific ports to be accessed on each VPS. I can control which ports any given VPS has open, allowing a web VPS to be accessed from anywhere on ports 80 and 443, but blocking a database VPS completely to the outside and only allowing a certain other VPS to access it. The configuration works well for my current needs. Note that there is not a hardware firewall protecting the virtualization hosts in place at this time. However, the KVM hosts only have port 22 open, are running nothing except KVM and SSH, and even port 22 cannot be accessed except for inside the netblock. I'm looking at possibly rethinking my network now that I have a client who needs to transition from a single VPS onto two dedicated servers (one web and one DB). A different customer already has a single dedicated server that is not behind any firewall except iptables running on the system. Should I require that each dedicated server customer have their own dedicated firewall? Or can I utilize a single network-wide firewall for multiple customer clusters? I'm familiar with iptables, and am currently thinking I'll use it for any firewalls/routers that I need. But I don't necessarily want to use up 1U of space in my rack for each firewall, nor the power consumption each firewall server will take. So I'm considering a hardware firewall. Any suggestions on what is a good approach?

    Read the article

< Previous Page | 58 59 60 61 62 63 64 65 66 67 68 69  | Next Page >