Re: cyrus perdtion , connections not dying
>> Usually we have around 700k simultaneous pop connections on the real >> servers now with perdition we have 3000+ connections > > Er, I didn't write that. That was Ram, the OP. Different person, different > institution, similar issues. We don't offer POP here. Ah oops, got a bit mixed up there. > We ran into the same problem, the performance limitation on our front end > cluster was (is) the number of processes. > Unfortunately, the same problem applies. Actually, it's worse. Cyrus > processes use more RAM: typically 2.0-3.5 MB per process as opposed to > 1.5 - 1.9 MB per process. I've even seen cyrus processes with up to 30MB. There seems to be some confusion here over if the problem is the size of the processes, or the actual number of processes. In many cases, the size of the process won't be the problem. In all UNIX OS's, the size the process is doesn't necessarily correlate very strongly to how much memory it's really using because lots of that can be shared between many processes (COW forking), or just mmaped files into the process space, which might not actually be all in memory at that moment. What's the actual memory status of your production machine (used/free/etc) and the cpu status of your production machine (idle/user/system/iowait)? Then you can work out what your real bottleneck is first. Rob Cyrus Home Page: http://cyrusimap.web.cmu.edu/ Cyrus Wiki/FAQ: http://cyrusimap.web.cmu.edu/twiki List Archives/Info: http://asg.web.cmu.edu/cyrus/mailing-list.html
Re: cyrus perdtion , connections not dying
On Tue, 8 Apr 2008, Ian Eiloart wrote: > --On 7 April 2008 10:20:33 -0700 Andrew Morgan <[EMAIL PROTECTED]> wrote: > >> On Mon, 7 Apr 2008, Ian Eiloart wrote: >> .. >>> >>> Is there a way to limit Cyrus process sizes at all? I guess I can take a >>> look at my compilation options to try to reduce the starting size, but >>> can I limit the process growth? >> >> How are you calculating the process size? I'm curious to compare it with >> my Cyrus installation. >> >> Andy > > I'm using the RSIZE reported in top. Here, I've got two users on a > development machine: > > top -U cyrus -o rsize > % top -U cyrus -o rsize -l 1 > Processes: 100 total, 4 running, 2 stuck, 94 sleeping... 292 threads > 11:27:10 > > Load Avg: 1.46, 1.38, 1.38CPU usage: 7.69% user, 46.15% sys, 46.15% > idle > SharedLibs: num =7, resident = 67M code, 1600K data, 5088K linkedit. > MemRegions: num = 9491, resident = 237M + 22M private, 124M shared. > PhysMem: 658M wired, 561M active, 28M inactive, 1247M used, 2799M free. > VM: 4439M + 372M 105234(0) pageins, 0(0) pageouts > > PID COMMAND %CPU TIME #TH #PRTS #MREGS RPRVT RSHRD RSIZE VSIZE > 71 master 0.0% 1:50:34 115+47 70M+ 2364K+ 70M+ 84M+ > 30219 proxyd 0.0% 0:01.78 115+ 105 672K+ 8168K+ 3552K+ 74M+ > 1076 proxyd 0.0% 0:00.83 115+ 110 524K+ 8176K+ 3388K+ 74M+ > 1093 proxyd 0.0% 0:02.56 115+ 110 516K+ 8176K+ 3332K+ 74M+ > 94 proxyd 0.0% 0:00.95 115+ 110 524K+ 8176K+ 3332K+ 74M+ > 35737 proxyd 0.0% 0:00.37 115+ 104 660K+ 8168K+ 3292K+ 74M+ > 7483 proxyd 0.0% 0:00.19 115+ 107 408K+ 8172K+ 2040K+ 74M+ > 84 proxyd 0.0% 0:00.02 115+ 107 408K+ 8172K+ 2040K+ 74M+ > 53828 proxyd 0.0% 0:00.27 115+ 101 464K+ 8164K+ 2024K+ 74M+ > 51026 proxyd 0.0% 0:00.27 115+ 101 464K+ 8164K+ 2024K+ 74M+ > 46021 proxyd 0.0% 0:00.27 115+ 101 464K+ 8164K+ 2024K+ 74M+ > 44932 proxyd 0.0% 0:00.23 115+ 101 464K+ 8164K+ 2024K+ 74M+ > 12886 proxyd 0.0% 0:00.20 115+ 101 464K+ 8164K+ 2024K+ 74M+ > 102 proxyd 0.0% 0:00.02 115+ 107 404K+ 8172K+ 2024K+ 74M+ > 100 proxyd 0.0% 0:00.02 115+ 107 404K+ 8172K+ 2024K+ 74M+ > 95 proxyd 0.0% 0:00.02 115+ 107 404K+ 8172K+ 2024K+ 74M+ > 93 proxyd 0.0% 0:00.02 115+ 107 404K+ 8172K+ 2024K+ 74M+ > 83 proxyd 0.0% 0:00.02 115+ 107 404K+ 8172K+ 2024K+ 74M+ > 57073 mupdate 0.0% 0:00.12 114+23 136K+ 1216K+ 604K+ 19M+ > 57074 mupdate 0.0% 0:00.05 113+23 128K+ 1216K+ 588K+ 19M+ > 88 launchd 0.0% 32:36.05 326+26 120K+ 296K+ 464K+ 18M+ These match what I see on my Cyrus Murder machines (linux 2.6) as well. With 328 proxyd processes currently running on a frontend, top reports: Tasks: 388 total, 1 running, 387 sleeping, 0 stopped, 0 zombie Cpu(s): 20.4%us, 1.0%sy, 0.0%ni, 78.3%id, 0.0%wa, 0.1%hi, 0.2%si, 0.0%st Mem: 2076724k total, 634260k used, 1442464k free, 162364k buffers Swap: 252k total,0k used, 252k free, 216200k cached Which seems to roughly agree with a 2MB per proxyd rule-of-thumb. Andy Cyrus Home Page: http://cyrusimap.web.cmu.edu/ Cyrus Wiki/FAQ: http://cyrusimap.web.cmu.edu/twiki List Archives/Info: http://asg.web.cmu.edu/cyrus/mailing-list.html
Re: cyrus perdtion , connections not dying
--On 8 April 2008 21:59:21 +1000 Rob Mueller <[EMAIL PROTECTED]> wrote: > Ok, so lets get some more details here. > >> We've got about 12,000 users, many of whom spend most of the day in >> lectures, and tend to read their email at lunch time. Some staff with >> large >> mail folders tend to stay logged on all day. > >> Currently we have four OSX 10.4 servers with 6GB of RAM each. We've >> purchased six OSX 10.5 machines with 4GB of RAM each, but worry that this >> might not be adequate. > > So you've currently got 4 servers. Are these servers running both cyrus > and perdition? Or are they on separate servers? Just perdition, for a (now completed) migration from UoW IMAP. Our new cluster, under development, will have a cyrus murder with separate front ends. > It seems you're running perdition and a murder setup so you can migrate > from an existing system. Is that right? Do you have separated frontends > and backends, or running them altogther? > >> Usually we have around 700k simultaneous pop connections on the real >> servers now with perdition we have 3000+ connections Er, I didn't write that. That was Ram, the OP. Different person, different institution, similar issues. We don't offer POP here. > I presume you menat 700 there. Is that across all your servers, or just > one of the four servers? That seems like quite a few POP connections. Do > you really mean POP connections, or do you mean IMAP? Or do you mean POP > and IMAP? How many "imapd" and "pop3d" processes do you see on each > server? How many do you see in total across all 4 servers? > > Rob > -- Ian Eiloart IT Services, University of Sussex x3148 Cyrus Home Page: http://cyrusimap.web.cmu.edu/ Cyrus Wiki/FAQ: http://cyrusimap.web.cmu.edu/twiki List Archives/Info: http://asg.web.cmu.edu/cyrus/mailing-list.html
Re: Permission denied on a MaillBox
ram ha scritto: > On Tue, 2008-04-08 at 12:17 +0200, Riccardo Riva wrote: >> Hi all, >> >> i have a Cyrus Server on which an admin had deleted the unix user >> (biinded to a mailbox) but not the mailbox. >> At this moment i'm not able to delete the mailbox using cyradm because >> when I'm trying to do dm user. I have the follow : >> >> deletemailbox: Permission denied >> >> If I try to set acl again on the mailbox, I'll have : >> >> setaclmailbox: cyrus: lrsiwpcda: System I/O error >> > On the file system the "cyrus" user must own the mailbox directory > if it is not chown it appropriately > > For eg , on my system( RHEL ) > chown cyrus:mail /var/spool/imap/domain/* Hi, thank you for your response. I've just checked file system permissions, and it's not a problem. Bye Ric Cyrus Home Page: http://cyrusimap.web.cmu.edu/ Cyrus Wiki/FAQ: http://cyrusimap.web.cmu.edu/twiki List Archives/Info: http://asg.web.cmu.edu/cyrus/mailing-list.html
Re: cyrus perdtion , connections not dying
Ok, so lets get some more details here. > We've got about 12,000 users, many of whom spend most of the day in > lectures, and tend to read their email at lunch time. Some staff with > large > mail folders tend to stay logged on all day. > Currently we have four OSX 10.4 servers with 6GB of RAM each. We've > purchased six OSX 10.5 machines with 4GB of RAM each, but worry that this > might not be adequate. So you've currently got 4 servers. Are these servers running both cyrus and perdition? Or are they on separate servers? It seems you're running perdition and a murder setup so you can migrate from an existing system. Is that right? Do you have separated frontends and backends, or running them altogther? > Usually we have around 700k simultaneous pop connections on the real > servers now with perdition we have 3000+ connections I presume you menat 700 there. Is that across all your servers, or just one of the four servers? That seems like quite a few POP connections. Do you really mean POP connections, or do you mean IMAP? Or do you mean POP and IMAP? How many "imapd" and "pop3d" processes do you see on each server? How many do you see in total across all 4 servers? Rob Cyrus Home Page: http://cyrusimap.web.cmu.edu/ Cyrus Wiki/FAQ: http://cyrusimap.web.cmu.edu/twiki List Archives/Info: http://asg.web.cmu.edu/cyrus/mailing-list.html
Re: Permission denied on a MaillBox
On Tue, 2008-04-08 at 12:17 +0200, Riccardo Riva wrote: > Hi all, > > i have a Cyrus Server on which an admin had deleted the unix user > (biinded to a mailbox) but not the mailbox. > At this moment i'm not able to delete the mailbox using cyradm because > when I'm trying to do dm user. I have the follow : > > deletemailbox: Permission denied > > If I try to set acl again on the mailbox, I'll have : > > setaclmailbox: cyrus: lrsiwpcda: System I/O error > On the file system the "cyrus" user must own the mailbox directory if it is not chown it appropriately For eg , on my system( RHEL ) chown cyrus:mail /var/spool/imap/domain/* > what can I do to delete that mailbox ? > > I've just even try to reconstruct to mailbox using > > /usr/lib/cyrus/bin/reconstruct -rfx "user.mailbox" > > but without any results. > > Thanks in advance > > Riccardo > > > Cyrus Home Page: http://cyrusimap.web.cmu.edu/ > Cyrus Wiki/FAQ: http://cyrusimap.web.cmu.edu/twiki > List Archives/Info: http://asg.web.cmu.edu/cyrus/mailing-list.html Cyrus Home Page: http://cyrusimap.web.cmu.edu/ Cyrus Wiki/FAQ: http://cyrusimap.web.cmu.edu/twiki List Archives/Info: http://asg.web.cmu.edu/cyrus/mailing-list.html
Re: cyrus perdtion , connections not dying
--On 8 April 2008 07:32:27 +1000 Rob Mueller <[EMAIL PROTECTED]> wrote: >> opposed to 1.5 - 1.9 MB per process. I've even seen cyrus processes with >> up >> to 30MB. > > cyrus uses mmap a lot. The processes probably aren't actually really > growing, they just look bigger because a user has a folder selected and > the cyrus.index file has been mmaped into the process space. If they > unselect the folder, the process size will drop back down again. > > How much RAM does your machine have and how many users are you trying to > support? We've got about 12,000 users, many of whom spend most of the day in lectures, and tend to read their email at lunch time. Some staff with large mail folders tend to stay logged on all day. Currently we have four OSX 10.4 servers with 6GB of RAM each. We've purchased six OSX 10.5 machines with 4GB of RAM each, but worry that this might not be adequate. I'm not sure about mmap. Is cyrus.index mmaped into process space on the proxy server? I currently have a 67MB mailbox with 15,000 messages open. I'm not seeing a particularly large process reflecting that fact. I suppose that's a good thing, though. > > As a point of reference. We have quite old machines with 8G RAM, 8T > attached storage that easily handle 2000+ processes under linux just fine. > > Rob > -- Ian Eiloart IT Services, University of Sussex x3148 Cyrus Home Page: http://cyrusimap.web.cmu.edu/ Cyrus Wiki/FAQ: http://cyrusimap.web.cmu.edu/twiki List Archives/Info: http://asg.web.cmu.edu/cyrus/mailing-list.html
Re: cyrus perdtion , connections not dying
--On 7 April 2008 10:20:33 -0700 Andrew Morgan <[EMAIL PROTECTED]> wrote: > On Mon, 7 Apr 2008, Ian Eiloart wrote: > .. >> >> Is there a way to limit Cyrus process sizes at all? I guess I can take a >> look at my compilation options to try to reduce the starting size, but >> can I limit the process growth? > > How are you calculating the process size? I'm curious to compare it with > my Cyrus installation. > > Andy I'm using the RSIZE reported in top. Here, I've got two users on a development machine: top -U cyrus -o rsize % top -U cyrus -o rsize -l 1 Processes: 100 total, 4 running, 2 stuck, 94 sleeping... 292 threads 11:27:10 Load Avg: 1.46, 1.38, 1.38CPU usage: 7.69% user, 46.15% sys, 46.15% idle SharedLibs: num =7, resident = 67M code, 1600K data, 5088K linkedit. MemRegions: num = 9491, resident = 237M + 22M private, 124M shared. PhysMem: 658M wired, 561M active, 28M inactive, 1247M used, 2799M free. VM: 4439M + 372M 105234(0) pageins, 0(0) pageouts PID COMMAND %CPU TIME #TH #PRTS #MREGS RPRVT RSHRD RSIZE VSIZE 71 master 0.0% 1:50:34 115+47 70M+ 2364K+ 70M+ 84M+ 30219 proxyd 0.0% 0:01.78 115+ 105 672K+ 8168K+ 3552K+ 74M+ 1076 proxyd 0.0% 0:00.83 115+ 110 524K+ 8176K+ 3388K+ 74M+ 1093 proxyd 0.0% 0:02.56 115+ 110 516K+ 8176K+ 3332K+ 74M+ 94 proxyd 0.0% 0:00.95 115+ 110 524K+ 8176K+ 3332K+ 74M+ 35737 proxyd 0.0% 0:00.37 115+ 104 660K+ 8168K+ 3292K+ 74M+ 7483 proxyd 0.0% 0:00.19 115+ 107 408K+ 8172K+ 2040K+ 74M+ 84 proxyd 0.0% 0:00.02 115+ 107 408K+ 8172K+ 2040K+ 74M+ 53828 proxyd 0.0% 0:00.27 115+ 101 464K+ 8164K+ 2024K+ 74M+ 51026 proxyd 0.0% 0:00.27 115+ 101 464K+ 8164K+ 2024K+ 74M+ 46021 proxyd 0.0% 0:00.27 115+ 101 464K+ 8164K+ 2024K+ 74M+ 44932 proxyd 0.0% 0:00.23 115+ 101 464K+ 8164K+ 2024K+ 74M+ 12886 proxyd 0.0% 0:00.20 115+ 101 464K+ 8164K+ 2024K+ 74M+ 102 proxyd 0.0% 0:00.02 115+ 107 404K+ 8172K+ 2024K+ 74M+ 100 proxyd 0.0% 0:00.02 115+ 107 404K+ 8172K+ 2024K+ 74M+ 95 proxyd 0.0% 0:00.02 115+ 107 404K+ 8172K+ 2024K+ 74M+ 93 proxyd 0.0% 0:00.02 115+ 107 404K+ 8172K+ 2024K+ 74M+ 83 proxyd 0.0% 0:00.02 115+ 107 404K+ 8172K+ 2024K+ 74M+ 57073 mupdate 0.0% 0:00.12 114+23 136K+ 1216K+ 604K+ 19M+ 57074 mupdate 0.0% 0:00.05 113+23 128K+ 1216K+ 588K+ 19M+ 88 launchd 0.0% 32:36.05 326+26 120K+ 296K+ 464K+ 18M+ And, these are the first few processes on a production server. There are actually 844 processes altogether, but it's a light day. % top -U nobody -o rsize -l 1 | grep perdition 10080 perdition0.0% 0:00.01 11556 348K 3.91M 1.67M 32.9M 10067 perdition0.0% 0:00.01 11556 348K 3.91M 1.67M 32.9M 10054 perdition0.0% 0:00.01 11556 348K 3.91M 1.67M 32.9M -- Ian Eiloart IT Services, University of Sussex x3148 Cyrus Home Page: http://cyrusimap.web.cmu.edu/ Cyrus Wiki/FAQ: http://cyrusimap.web.cmu.edu/twiki List Archives/Info: http://asg.web.cmu.edu/cyrus/mailing-list.html
Re: can not run cyradm from remote server .. realm gets appended to username
On Mon, 2008-04-07 at 14:01 +0200, Alain Spineux wrote: > On Thu, Apr 3, 2008 at 11:33 PM, ram <[EMAIL PROTECTED]> wrote: > > On my cryus imap server , when I try login from my localserver cyrus > > login works > > You used localhost ? Only localhost is working then. > Try using the IP address you used from your remote server, it should fail too. > > > > > > > But when I login from a remote server realm gets added autmatically , > > causing login to fail > > man imad.conf, look for "virtdomains" option, it use the ip address of > your interface to guess a > realm. The key is arround options virtualdomains, defaultdomains and > loginrealms, > but it depend of what you want ! Dou you have multiple domains ? > Yes I have multiple domains. I need support for virtualdomains Actually I solved the issue , using a silly workaround. I had another interface on the remote cyrus server and when I cyradm to the secondary interface IP, No realm gets attached in SASL. No idea why it works , but it sure does !! > > > > > > In my sasl logs I can see for a local connection > > - > > saslauthd[21947] :do_auth : auth success: [user=cyrus] > > [service=imap] [realm=] [mech=pam] > > saslauthd[21947] :do_request : response: OK > > -- > > > > > > > > > > > > For a remote connection > > - > > saslauthd[24011] :do_auth : auth failure: > > [EMAIL PROTECTED] [service=imap] [realm=netcore.co.in] > > [mech=pam] [reason=PAM auth error] > > -- > > > > > > What can be seen simply is that the realm netcore.co.in is getting added > > automatically. How can I avoid this > > > > > > Cyrus Home Page: http://cyrusimap.web.cmu.edu/ > > Cyrus Wiki/FAQ: http://cyrusimap.web.cmu.edu/twiki > > List Archives/Info: http://asg.web.cmu.edu/cyrus/mailing-list.html > > > > > Cyrus Home Page: http://cyrusimap.web.cmu.edu/ Cyrus Wiki/FAQ: http://cyrusimap.web.cmu.edu/twiki List Archives/Info: http://asg.web.cmu.edu/cyrus/mailing-list.html
Permission denied on a MaillBox
Hi all, i have a Cyrus Server on which an admin had deleted the unix user (biinded to a mailbox) but not the mailbox. At this moment i'm not able to delete the mailbox using cyradm because when I'm trying to do dm user. I have the follow : deletemailbox: Permission denied If I try to set acl again on the mailbox, I'll have : setaclmailbox: cyrus: lrsiwpcda: System I/O error what can I do to delete that mailbox ? I've just even try to reconstruct to mailbox using /usr/lib/cyrus/bin/reconstruct -rfx "user.mailbox" but without any results. Thanks in advance Riccardo Cyrus Home Page: http://cyrusimap.web.cmu.edu/ Cyrus Wiki/FAQ: http://cyrusimap.web.cmu.edu/twiki List Archives/Info: http://asg.web.cmu.edu/cyrus/mailing-list.html