<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <font face="monospace">Better don't :), as we never explored all the
      possible side effects of such combination of versions....<br>
      <br>
      Regards,<br>
    </font>
    <pre class="moz-signature" cols="72">Bogdan-Andrei Iancu

OpenSIPS Founder and Developer
  <a class="moz-txt-link-freetext" href="https://www.opensips-solutions.com">https://www.opensips-solutions.com</a>
OpenSIPS eBootcamp 2021 
  <a class="moz-txt-link-freetext" href="https://opensips.org/training/OpenSIPS_eBootcamp_2021/">https://opensips.org/training/OpenSIPS_eBootcamp_2021/</a></pre>
    <div class="moz-cite-prefix">On 1/19/22 9:45 AM, Schneur Rosenberg
      wrote:<br>
    </div>
    <blockquote type="cite"
cite="mid:CANvjR0WvsY0uSBHRKzbenA0KpUHbB_AvsAa37NSk3C_mH2k28w@mail.gmail.com">
      <meta http-equiv="content-type" content="text/html; charset=UTF-8">
      <div dir="auto">That's a given the question is can it mess up the
        active server if the passive one was updated already? I'm not
        planning on leaving it like that I just like to do the upgrades
        slowly, one at a time and test it and only then to upgrade the
        second one.
        <div dir="auto"><br>
        </div>
        <div dir="auto">Scott</div>
      </div>
      <br>
      <div class="gmail_quote">
        <div dir="ltr" class="gmail_attr">On Wed, Jan 19, 2022, 09:26
          Bogdan-Andrei Iancu <<a href="mailto:bogdan@opensips.org"
            moz-do-not-send="true">bogdan@opensips.org</a>> wrote:<br>
        </div>
        <blockquote class="gmail_quote" style="margin:0 0 0
          .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi Schneur,<br>
          <br>
          It is strongly recommend that all OpenSIPS nodes in a cluster 
          to have <br>
          the same version.<br>
          <br>
          Best regards,<br>
          <br>
          Bogdan-Andrei Iancu<br>
          <br>
          OpenSIPS Founder and Developer<br>
             <a href="https://www.opensips-solutions.com"
            rel="noreferrer noreferrer" target="_blank"
            moz-do-not-send="true">https://www.opensips-solutions.com</a><br>
          OpenSIPS eBootcamp 2021<br>
             <a
            href="https://opensips.org/training/OpenSIPS_eBootcamp_2021/"
            rel="noreferrer noreferrer" target="_blank"
            moz-do-not-send="true">https://opensips.org/training/OpenSIPS_eBootcamp_2021/</a><br>
          <br>
          On 1/18/22 6:08 PM, Schneur Rosenberg wrote:<br>
          > Hi, it seems like it was fixed in 3.2, I will have to
          migrate all my<br>
          > servers, I use binary replication will it break if one
          server is<br>
          > running 2.4 and the other 3.2? its a active/passive setup
          so I will<br>
          > take down one at a time and upgrade it, I'm just worried
          what will<br>
          > happen while one is 3.2 and the second one is still 2.4,
          in the past I<br>
          > disabled the replication while I was doing the updates
          and I'm<br>
          > wondering if its necessary.<br>
          ><br>
          > thanks<br>
          > Scott (Schneur)<br>
          ><br>
          > On Fri, Dec 17, 2021 at 6:21 PM Bogdan-Andrei Iancu <<a
            href="mailto:bogdan@opensips.org" target="_blank"
            rel="noreferrer" moz-do-not-send="true">bogdan@opensips.org</a>>
          wrote:<br>
          >> While trying to reproduce (as I failed to do so), I
          noticed you mentioned this is on version 2.4.11, right ? As I
          was testing on 3.2 without getting the leak.<br>
          >><br>
          >> Could you try on 3.2/3.1 ? Keep in mind 2.4 is not
          maintained anymore :(<br>
          >><br>
          >> Regards,<br>
          >><br>
          >> Bogdan-Andrei Iancu<br>
          >><br>
          >> OpenSIPS Founder and Developer<br>
          >>    <a href="https://www.opensips-solutions.com"
            rel="noreferrer noreferrer" target="_blank"
            moz-do-not-send="true">https://www.opensips-solutions.com</a><br>
          >> OpenSIPS eBootcamp 2021<br>
          >>    <a
            href="https://opensips.org/training/OpenSIPS_eBootcamp_2021/"
            rel="noreferrer noreferrer" target="_blank"
            moz-do-not-send="true">https://opensips.org/training/OpenSIPS_eBootcamp_2021/</a><br>
          >><br>
          >> On 12/17/21 1:40 PM, Schneur Rosenberg wrote:<br>
          >><br>
          >> Thanks Bogdan!, this is my entire local_route, all my
          dst_uri's are IP only.<br>
          >><br>
          >> On Fri, Dec 17, 2021, 12:36 Bogdan-Andrei Iancu <<a
            href="mailto:bogdan@opensips.org" target="_blank"
            rel="noreferrer" moz-do-not-send="true">bogdan@opensips.org</a>>
          wrote:<br>
          >>> Hi Schneur,<br>
          >>><br>
          >>> I suspect that the leaking mk_proxy is related to
          the changing of the<br>
          >>> RURI in local route. Let me test your snippet.
          BTW, is that the whole<br>
          >>> processing you do in local route? is the $rd
          (from LB) a FQDN or<br>
          >>> straight IP ?<br>
          >>><br>
          >>> Regards,<br>
          >>><br>
          >>> Bogdan-Andrei Iancu<br>
          >>><br>
          >>> OpenSIPS Founder and Developer<br>
          >>>     <a href="https://www.opensips-solutions.com"
            rel="noreferrer noreferrer" target="_blank"
            moz-do-not-send="true">https://www.opensips-solutions.com</a><br>
          >>> OpenSIPS eBootcamp 2021<br>
          >>>     <a
            href="https://opensips.org/training/OpenSIPS_eBootcamp_2021/"
            rel="noreferrer noreferrer" target="_blank"
            moz-do-not-send="true">https://opensips.org/training/OpenSIPS_eBootcamp_2021/</a><br>
          >>><br>
          >>> On 12/16/21 9:43 AM, Schneur Rosenberg wrote:<br>
          >>>> Hi Bogdan<br>
          >>>><br>
          >>>> I think I found the issue, I recently added
          these lines of code,<br>
          >>>> because of a probing issue I was having, I
          just searched from my<br>
          >>>> previous tickets and I see that you have
          warned me about the<br>
          >>>> implications but for some reason I never read
          the message.<br>
          >>>><br>
          >>>> Here is the old ticket<br>
          >>>> <a
href="https://www.mail-archive.com/users@lists.opensips.org/msg43301.html"
            rel="noreferrer noreferrer" target="_blank"
            moz-do-not-send="true">https://www.mail-archive.com/users@lists.opensips.org/msg43301.html</a><br>
          >>>> the reason I'm using INVITE to probe is
          because I want the servers<br>
          >>>> that were probed not only to respond but also
          check if the database is<br>
          >>>> working, I did it this way because I had
          cases where mysql crashed but<br>
          >>>> my asterisk servers were still responding to
          the probe but all of the<br>
          >>>> calls just hung, so I do a invite and it does
          a DB lookup and it will<br>
          >>>> only return a positive message if it was able
          to query the DB, do you<br>
          >>>> have a better solution? at the time I set it
          up I couldn't run a query<br>
          >>>> on receipt of a OPTIONS but perhaps I didn't
          look good enough :-),<br>
          >>>> either way can I do anything to make sure
          this code doesn't leak<br>
          >>>> memory? this probing has worked for years
          until I needed the Contact<br>
          >>>> header.<br>
          >>>><br>
          >>>> local_route {<br>
          >>>>        if (is_method("INVITE")&&
          $fU=="pingTest"){<br>
          >>>>           $ru=<a class="moz-txt-link-rfc2396E" href="sip:s@">"sip:s@"</a>+$rd ;<br>
          >>>>           append_hf("Contact:
          <a class="moz-txt-link-rfc2396E" href="sip:pingTest@$fd:5060"><sip:pingTest@$fd:5060></a>\r\n");<br>
          >>>>           exit;<br>
          >>>>        }<br>
          >>>> }<br>
          >>>><br>
          >>>> On Fri, Dec 10, 2021 at 2:16 PM Schneur
          Rosenberg<br>
          >>>> <<a href="mailto:rosenberg11219@gmail.com"
            target="_blank" rel="noreferrer" moz-do-not-send="true">rosenberg11219@gmail.com</a>>
          wrote:<br>
          >>>>> Hi Bogdan,<br>
          >>>>><br>
          >>>>> I did it on a backup server, its also
          leaking memory but at a slower<br>
          >>>>> pace, I'm attaching the logs when running
          kill -SIGUSR1 on the pid<br>
          >>>>> that's growing in size, it still has
          available memory, I hop this will<br>
          >>>>> give you a clue.<br>
          >>>>><br>
          >>>>> Here is a pastbin to the loggs <a
            href="https://pastebin.com/KJVb9Y75" rel="noreferrer
            noreferrer" target="_blank" moz-do-not-send="true">https://pastebin.com/KJVb9Y75</a><br>
          >>>>><br>
          >>>>> On Fri, Dec 10, 2021 at 11:00 AM Schneur
          Rosenberg<br>
          >>>>> <<a
            href="mailto:rosenberg11219@gmail.com" target="_blank"
            rel="noreferrer" moz-do-not-send="true">rosenberg11219@gmail.com</a>>
          wrote:<br>
          >>>>>> Thank you, does this reduce
          performance? can I leave it enabled on a<br>
          >>>>>> production machine? I will wait for
          the memory leak to be apparent and<br>
          >>>>>> I'll post the result.<br>
          >>>>>><br>
          >>>>>><br>
          >>>>>> On Thu, Dec 9, 2021 at 12:31 PM
          Bogdan-Andrei Iancu <<a href="mailto:bogdan@opensips.org"
            target="_blank" rel="noreferrer" moz-do-not-send="true">bogdan@opensips.org</a>>
          wrote:<br>
          >>>>>>> Hi Schneur,<br>
          >>>>>>><br>
          >>>>>>> Just follow the<br>
          >>>>>>> <a
            href="https://www.opensips.org/Documentation/TroubleShooting-OutOfMem"
            rel="noreferrer noreferrer" target="_blank"
            moz-do-not-send="true">https://www.opensips.org/Documentation/TroubleShooting-OutOfMem</a>
          and<br>
          >>>>>>> provide the dump. This is the
          only way to investigate this.<br>
          >>>>>>><br>
          >>>>>>> Regards,<br>
          >>>>>>><br>
          >>>>>>> Bogdan-Andrei Iancu<br>
          >>>>>>><br>
          >>>>>>> OpenSIPS Founder and Developer<br>
          >>>>>>>      <a
            href="https://www.opensips-solutions.com" rel="noreferrer
            noreferrer" target="_blank" moz-do-not-send="true">https://www.opensips-solutions.com</a><br>
          >>>>>>> OpenSIPS eBootcamp 2021<br>
          >>>>>>>      <a
            href="https://opensips.org/training/OpenSIPS_eBootcamp_2021/"
            rel="noreferrer noreferrer" target="_blank"
            moz-do-not-send="true">https://opensips.org/training/OpenSIPS_eBootcamp_2021/</a><br>
          >>>>>>><br>
          >>>>>>> On 12/8/21 12:14 PM, Schneur
          Rosenberg wrote:<br>
          >>>>>>>> I just noticed that process
          88 runs the timer handler, perhaps this<br>
          >>>>>>>> might shed light on whats
          going on.<br>
          >>>>>>>><br>
          >>>>>>>> opensipsctl fifo ps<br>
          >>>>>>>> Process::  ID=88 PID=5327
          Type=Timer handler<br>
          >>>>>>>><br>
          >>>>>>>> On Wed, Dec 8, 2021 at 10:55
          AM Schneur Rosenberg<br>
          >>>>>>>> <<a
            href="mailto:rosenberg11219@gmail.com" target="_blank"
            rel="noreferrer" moz-do-not-send="true">rosenberg11219@gmail.com</a>>
          wrote:<br>
          >>>>>>>>> Now a few hours later
          this is what I'm getting<br>
          >>>>>>>>> Dec  8 09:50:13
          /sbin/opensips[21699]: ERROR:nathelper:nh_timer: out<br>
          >>>>>>>>> of pkg memory<br>
          >>>>>>>>> Dec  8 09:50:16
          /sbin/opensips[21699]: WARNING:core:fm_malloc: not<br>
          >>>>>>>>> enough continuous free
          pkg memory (3024 bytes left, need 5128),<br>
          >>>>>>>>> attempting
          defragmentation... please increase the "-M" command line<br>
          >>>>>>>>> parameter!<br>
          >>>>>>>>> Dec  8 09:50:16
          /sbin/opensips[21699]: ERROR:core:fm_malloc: not<br>
          >>>>>>>>> enough free pkg memory
          (3024 bytes left, need 5128), please increase<br>
          >>>>>>>>> the "-M" command line
          parameter!<br>
          >>>>>>>>><br>
          >>>>>>>>> Here is the last 20
          package memory max_used_size<br>
          >>>>>>>>> pkmem:70-max_used_size::
          1009584<br>
          >>>>>>>>> pkmem:71-max_used_size::
          1009584<br>
          >>>>>>>>> pkmem:72-max_used_size::
          1009584<br>
          >>>>>>>>> pkmem:73-max_used_size::
          1009584<br>
          >>>>>>>>> pkmem:74-max_used_size::
          1009584<br>
          >>>>>>>>> pkmem:75-max_used_size::
          1009584<br>
          >>>>>>>>> pkmem:76-max_used_size::
          1009584<br>
          >>>>>>>>> pkmem:77-max_used_size::
          1009584<br>
          >>>>>>>>> pkmem:78-max_used_size::
          1009584<br>
          >>>>>>>>> pkmem:79-max_used_size::
          1009584<br>
          >>>>>>>>> pkmem:80-max_used_size::
          1044752<br>
          >>>>>>>>> pkmem:81-max_used_size::
          1075552<br>
          >>>>>>>>> pkmem:82-max_used_size::
          1116848<br>
          >>>>>>>>> pkmem:83-max_used_size::
          1117456<br>
          >>>>>>>>> pkmem:84-max_used_size::
          1102640<br>
          >>>>>>>>> pkmem:85-max_used_size::
          1306992<br>
          >>>>>>>>> pkmem:86-max_used_size::
          1706304<br>
          >>>>>>>>> pkmem:87-max_used_size::
          2507000<br>
          >>>>>>>>> pkmem:88-max_used_size::
          4194264<br>
          >>>>>>>>> pkmem:89-max_used_size::
          1009584<br>
          >>>>>>>>><br>
          >>>>>>>>> And here is the real used
          size, you can see that process 88 maxed out<br>
          >>>>>>>>> pkmem:69-real_used_size::
          975528<br>
          >>>>>>>>> pkmem:70-real_used_size::
          978016<br>
          >>>>>>>>> pkmem:71-real_used_size::
          989592<br>
          >>>>>>>>> pkmem:72-real_used_size::
          951416<br>
          >>>>>>>>> pkmem:73-real_used_size::
          982496<br>
          >>>>>>>>> pkmem:74-real_used_size::
          965744<br>
          >>>>>>>>> pkmem:75-real_used_size::
          959424<br>
          >>>>>>>>> pkmem:76-real_used_size::
          949472<br>
          >>>>>>>>> pkmem:77-real_used_size::
          983080<br>
          >>>>>>>>> pkmem:78-real_used_size::
          961400<br>
          >>>>>>>>> pkmem:79-real_used_size::
          977808<br>
          >>>>>>>>> pkmem:80-real_used_size::
          978928<br>
          >>>>>>>>> pkmem:81-real_used_size::
          1009936<br>
          >>>>>>>>> pkmem:82-real_used_size::
          1110760<br>
          >>>>>>>>> pkmem:83-real_used_size::
          1116720<br>
          >>>>>>>>> pkmem:84-real_used_size::
          1096568<br>
          >>>>>>>>> pkmem:85-real_used_size::
          1300592<br>
          >>>>>>>>> pkmem:86-real_used_size::
          1699648<br>
          >>>>>>>>> pkmem:87-real_used_size::
          2501096<br>
          >>>>>>>>> pkmem:88-real_used_size::
          4191280<br>
          >>>>>>>>> pkmem:89-real_used_size::
          882528<br>
          >>>>>>>>><br>
          >>>>>>>>> On Tue, Dec 7, 2021 at
          7:53 PM Schneur Rosenberg<br>
          >>>>>>>>> <<a
            href="mailto:rosenberg11219@gmail.com" target="_blank"
            rel="noreferrer" moz-do-not-send="true">rosenberg11219@gmail.com</a>>
          wrote:<br>
          >>>>>>>>>> Hi, lately I'm
          getting  these errors in my logs.<br>
          >>>>>>>>>><br>
          >>>>>>>>>> ERROR:core:fm_malloc:
          not enough free pkg memory (1792 bytes left,<br>
          >>>>>>>>>> need 2184), please
          increase the "-M" command line para<br>
          >>>>>>>>>> meter!<br>
          >>>>>>>>>><br>
          >>>>>>>>>>
          CRITICAL:core:hostent_cpy: pkg memory allocation failure<br>
          >>>>>>>>>><br>
          >>>>>>>>>>
          ERROR:nathelper:nh_timer: out of pkg memory<br>
          >>>>>>>>>><br>
          >>>>>>>>>> ERROR:core:fm_malloc:
          not enough free pkg memory (5952 bytes left,<br>
          >>>>>>>>>> need 5408), please
          increase the "-M" command line para<br>
          >>>>>>>>>> meter!<br>
          >>>>>>>>>><br>
          >>>>>>>>>> I was on version
          2.4.8 and I upgraded to 2.4.11 and I'm monitoring the<br>
          >>>>>>>>>> max_used_size of the
          package memory, a few hours later I see that 2<br>
          >>>>>>>>>> processes keep on
          getting bigger, so far the rest are pretty stable, I<br>
          >>>>>>>>>> have 90 processes and
          87 and 88 are growing.<br>
          >>>>>>>>>><br>
          >>>>>>>>>> here you can see the
          last few processes, OpenSIPS set aside 4 mb per process.<br>
          >>>>>>>>>><br>
          >>>>>>>>>>
          pkmem:80-max_used_size:: 1009584<br>
          >>>>>>>>>>
          pkmem:81-max_used_size:: 1009584<br>
          >>>>>>>>>>
          pkmem:82-max_used_size:: 1009584<br>
          >>>>>>>>>>
          pkmem:83-max_used_size:: 1009584<br>
          >>>>>>>>>>
          pkmem:84-max_used_size:: 1009584<br>
          >>>>>>>>>>
          pkmem:85-max_used_size:: 1009584<br>
          >>>>>>>>>>
          pkmem:86-max_used_size:: 1143608<br>
          >>>>>>>>>>
          pkmem:87-max_used_size:: 1323256<br>
          >>>>>>>>>>
          pkmem:88-max_used_size:: 1831928<br>
          >>>>>>>>>>
          pkmem:89-max_used_size:: 1009584<br>
          >>>>>>>>>><br>
          >>>>>>>>>> Any hints where to
          start looking besides the solutions fund here.<br>
          >>>>>>>>>><br>
          >>>>>>>>>> <a
            href="https://www.opensips.org/Documentation/TroubleShooting-OutOfMem"
            rel="noreferrer noreferrer" target="_blank"
            moz-do-not-send="true">https://www.opensips.org/Documentation/TroubleShooting-OutOfMem</a><br>
          >>>>>>>>>><br>
          >>>>>>>>>> thank you<br>
          >>>>>>>>>> Scott<br>
          >>>>>>>>
          _______________________________________________<br>
          >>>>>>>> Users mailing list<br>
          >>>>>>>> <a
            href="mailto:Users@lists.opensips.org" target="_blank"
            rel="noreferrer" moz-do-not-send="true">Users@lists.opensips.org</a><br>
          >>>>>>>> <a
            href="http://lists.opensips.org/cgi-bin/mailman/listinfo/users"
            rel="noreferrer noreferrer" target="_blank"
            moz-do-not-send="true">http://lists.opensips.org/cgi-bin/mailman/listinfo/users</a><br>
          <br>
        </blockquote>
      </div>
    </blockquote>
    <br>
  </body>
</html>