[SOGo] BTS activities for Friday, January 05 2018
Title: BTS activities for Friday, January 05 2018 BTS Activities Home page: http://www.sogo.nu/bugs Project: SOGo For the period covering: Friday, January 05 2018 idlast updatestatus (resolution)categorysummary 4370 2018-01-05 08:44:56 resolved (fixed) i18n Wrong Lithuanian translation of weekdays short names -- users@sogo.nuhttps://inverse.ca/sogo/lists
RE: [SOGo] Performance Issues
As mentioned, when PREFORK is greater than 500, the service will not start. My server CPU average is about 5% and memory is about 60% with 0.4% swap. I can easily add more memory, but performance wise, we are barely touching the allocated specifications. Regards, Karl From: users-requ...@sogo.nu [mailto:users-requ...@sogo.nu] On Behalf Of Bruno CAPELETO Sent: Friday, January 5, 2018 11:36 AM To: users@sogo.nu Subject: Re: [SOGo] Performance Issues As far as I know, you need one sogo prefork for every single EAS client. Every sogo prefork needs some ram... 2018-01-05 16:26 GMT+01:00 Karl Hoffner : We have a single instance iRedmail/iRedAdmin-Pro/MySQL/SOGo instance running on CentOS. We initially started the server with about 40-50 mailboxes and users and had little or no issues. We recently added about 1500 mailboxes with 1500 users. All users will be using EAS - thus the need for SOGo and ActiveSync. Since the addition of the users we cannot use the SOGo web interface or iRedAdmin-Pro unless we restart the sogod service. This is required about every couple minutes. Sometimes as little as 60 seconds and in off hours, perhaps up to 5-6 minutes.. I did create a restart fiunction to restart the service every 5 minutes. This was not working as we still were unable to access. So we had it down to 2 minutes. This helped, but if you were sending message at that time, it would fail and you would have to rewrite. We have made changes to the PREFORK (initial default at 3), increase to 30, then increased to 50, then 100; now at 500. (Larger numbers than 500 prevent sogod from starting) We also had to up the SQL max_connections to 700 to function. We increased the WOWorkersCount to 500 then to 1800 (1500 EAS + 20%) We tried moving EAS to a new port and spawning its own process using the instructions in the following link; the new service will not start indicating the port is in use. https://sogo.nu/nc/support/faq/article/dedicated-separate-sogo-instance-for-activesync.html Using ps -A, only one instance is using 127.0.0.1:2 and none on 3 as per the instructions. Apparently the GNUstepDefaults file is not being read?? Anyway, this failed and we still basically have an unusable email server. Any help would be welcome as we requested professional support from SOGo with no contact back. If we cannot resolve, we will be forced to find an alternative as the performance for this many users is unacceptable. I was and am hoping to not have to waste the time to recreate a new mail server. CentOS: 7.36.1611 w/8 CPU and 16 GB memory iRedmail 0.9.6 MARIADB edition iRedAdmin-Pro 2.6.0 Regards, Karl CONFIDENTIALITY NOTICE: This electronic mail transmission and any documents, files, or previous e-mail messages attached to it may contain confidential information and are solely for the use of the individual or entity to which they were intended. If you have received this message in error, please notify the sender immediately by e-mail, discard any paper copies, and delete all electronic files of the message. -- users@sogo.nu https://inverse.ca/sogo/lists -- users@sogo.nu https://inverse.ca/sogo/lists -- users@sogo.nu https://inverse.ca/sogo/lists
Re: [SOGo] Performance Issues
As far as I know, you need one sogo prefork for every single EAS client. Every sogo prefork needs some ram... 2018-01-05 16:26 GMT+01:00 Karl Hoffner : > We have a single instance iRedmail/iRedAdmin-Pro/MySQL/SOGo instance > running on CentOS. We initially started the server with about 40-50 > mailboxes and users and had little or no issues. We recently added about > 1500 mailboxes with 1500 users. All users will be using EAS - thus the need > for SOGo and ActiveSync. Since the addition of the users we cannot use the > SOGo web interface or iRedAdmin-Pro unless we restart the sogod service. > This is required about every couple minutes. Sometimes as little as 60 > seconds and in off hours, perhaps up to 5-6 minutes.. I did create a > restart fiunction to restart the service every 5 minutes. This was not > working as we still were unable to access. So we had it down to 2 minutes. > This helped, but if you were sending message at that time, it would fail > and you would have to rewrite. > > We have made changes to the PREFORK (initial default at 3), increase to > 30, then increased to 50, then 100; now at 500. (Larger numbers than 500 > prevent sogod from starting) > We also had to up the SQL max_connections to 700 to function. > We increased the WOWorkersCount to 500 then to 1800 (1500 EAS + 20%) > We tried moving EAS to a new port and spawning its own process using the > instructions in the following link; the new service will not start > indicating the port is in use. > https://sogo.nu/nc/support/faq/article/dedicated- > separate-sogo-instance-for-activesync.html > Using ps -A, only one instance is using 127.0.0.1:2 and none on 3 > as per the instructions. Apparently the GNUstepDefaults file is not being > read?? Anyway, this failed and we still basically have an unusable email > server. > > Any help would be welcome as we requested professional support from SOGo > with no contact back. If we cannot resolve, we will be forced to find an > alternative as the performance for this many users is unacceptable. I was > and am hoping to not have to waste the time to recreate a new mail server. > > CentOS: 7.36.1611 w/8 CPU and 16 GB memory > iRedmail 0.9.6 MARIADB edition > iRedAdmin-Pro 2.6.0 > > Regards, > Karl > > > CONFIDENTIALITY NOTICE: This electronic mail transmission and any > documents, files, or previous e-mail messages attached to it may contain > confidential information and are solely for the use of the individual or > entity to which they were intended. If you have received this message in > error, please notify the sender immediately by e-mail, discard any paper > copies, and delete all electronic files of the message. > -- > users@sogo.nu > https://inverse.ca/sogo/lists -- users@sogo.nu https://inverse.ca/sogo/lists
Re: [SOGo] v2 development roadmap
Il 04/01/2018 20:48, Chris (chris2...@postbox.xyz) ha scritto: > On Mon, 18 Dec 2017 20:48:00 +0100 > lists wrote: > >> On 18-12-2017 16:46, Ludovic Marcotte (lmarco...@inverse.ca) wrote: >>> We plan to maintain it for many months to come. >> Just to chime in: We're happy to see this statement. >> >> We're not on v3, and our users would be very disappointed to see v2 >> go. > Same here. We really appreciate it. Is there by any chance that SOGo 2 will ever support the "single table" database structure as in v3? I currently run both 2 and 3, though having a simpler db structure would make any maintainance easier to me. Buon lavoro. /Alessandro Briosi/ *METAL.it Nord S.r.l.* Via Maioliche 57/C - 38068 Rovereto (TN) Tel.+39.0464.430130 - Fax +39.0464.437393 www.metalit.com -- users@sogo.nu https://inverse.ca/sogo/lists
[SOGo] Performance Issues
We have a single instance iRedmail/iRedAdmin-Pro/MySQL/SOGo instance running on CentOS. We initially started the server with about 40-50 mailboxes and users and had little or no issues. We recently added about 1500 mailboxes with 1500 users. All users will be using EAS - thus the need for SOGo and ActiveSync. Since the addition of the users we cannot use the SOGo web interface or iRedAdmin-Pro unless we restart the sogod service. This is required about every couple minutes. Sometimes as little as 60 seconds and in off hours, perhaps up to 5-6 minutes.. I did create a restart fiunction to restart the service every 5 minutes. This was not working as we still were unable to access. So we had it down to 2 minutes. This helped, but if you were sending message at that time, it would fail and you would have to rewrite. We have made changes to the PREFORK (initial default at 3), increase to 30, then increased to 50, then 100; now at 500. (Larger numbers than 500 prevent sogod from starting) We also had to up the SQL max_connections to 700 to function. We increased the WOWorkersCount to 500 then to 1800 (1500 EAS + 20%) We tried moving EAS to a new port and spawning its own process using the instructions in the following link; the new service will not start indicating the port is in use. https://sogo.nu/nc/support/faq/article/dedicated-separate-sogo-instance-for-activesync.html Using ps -A, only one instance is using 127.0.0.1:2 and none on 3 as per the instructions. Apparently the GNUstepDefaults file is not being read?? Anyway, this failed and we still basically have an unusable email server. Any help would be welcome as we requested professional support from SOGo with no contact back. If we cannot resolve, we will be forced to find an alternative as the performance for this many users is unacceptable. I was and am hoping to not have to waste the time to recreate a new mail server. CentOS: 7.36.1611 w/8 CPU and 16 GB memory iRedmail 0.9.6 MARIADB edition iRedAdmin-Pro 2.6.0 Regards, Karl CONFIDENTIALITY NOTICE: This electronic mail transmission and any documents, files, or previous e-mail messages attached to it may contain confidential information and are solely for the use of the individual or entity to which they were intended. If you have received this message in error, please notify the sender immediately by e-mail, discard any paper copies, and delete all electronic files of the message. -- users@sogo.nu https://inverse.ca/sogo/lists
[SOGo] Invitation reminders pop up - Calendar is already removed
Hi, we have the problem that for a group of users the invitation for an appointment pops up regularly, although the calendar is deleted. How can we stop the appointment showing up? Regards . Götz smime.p7s Description: S/MIME cryptographic signature
Re: [SOGo] 2 sogo servers (master/master)
Am 28.12.2017 um 19:37 schrieb Eric DUVAL (edu...@ligfy.fr): > Hello, > > I use SOGo3 on 2 servers (master/master). > DNS configured in roundrobin. (webmail.example.com with 2 IPs). > In front of SOGo servers I have 2 haproxy. Do you use 2 IPs in order to loadbalance between the two haproxys? Else I would use one service IP and switch that between the haproxys in case of failure/updates/replacement e.g. with heartbeat. Beware: There are still clients out there, who can not handle 2 IPs and will therefore only use the first one all the time. > Mencached is used on each SOGo server. > Yes, that is the fastest solution. > Is it ok to share the mysql database between the 2 SOGo servers? > Yes, of course, just make sure it can handle the load. Else you would need a synchronisation between two mysql servers (cluster mode). > I have a cache problem on the 2 servers: "[SOGoCache]> an error occurred > when caching value for key ... "SERVER HAS FAILED AND IS DISABLED UNTIL > TIMED RETRY" Do you have an idea to fix this problem? > Does memcached run at all? Which memcached settings do you have in sogo.conf? > What are the best practices for using 2 SOGo servers (master / master)? > As you can not predict, which of the servers the user uses, you have to share temporary Information between the two server e.g. via NFS too. By "temporary Information" I mean the content of SOGoMailSpoolPath and NGMimeBuildMimeTempDirectory in sogo.conf. The first is used for storing new emails and their attachments in order to transcode them into MIME format. The second is used to temporarily store file uploads till they are completed. Alternative would be to use source IP pinning, which is more complicated. Kind regards, Christian Mack -- Christian Mack Universität Konstanz Kommunikations-, Informations-, Medienzentrum (KIM) Abteilung Basisdienste 78457 Konstanz +49 7531 88-4416 smime.p7s Description: S/MIME Cryptographic Signature