Re: [systemd-devel] Starting units when a port is available for connections
> On 27 May 2015, at 8:40 pm, Andrei Borzenkov wrote: > > Hmm ... this sounds suspiciously like what D-Bus does. Did you consider > using D-Bus in your application? > > But for now there is no way to express such dependency in systemd; > D-Bus being exception, you can make services dependent on D-Bus end > points. I’ve considered it now :) I communicate with systemd via D-Bus for starting & stopping services. How does one write a service unit that depends on a D-Bus endpoint? Is this supported by systemd, or is this an application level thing? I’m unable to find anything in the systemd docs about creating dependencies on D-Bus endpoints. > I wonder - can your master service trigger startup of clients when it is > ready? Note that it can be done in completely generic way - it can > simply run something like cassandra.target and you can plug in any > client into this target. This is another option that I’ll play with. Thanks, Adam ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/systemd-devel
[systemd-devel] Starting units when a port is available for connections
Hi list, I’m running Cassandra (C*, a clustered database) as a systemd service. Currently this is just a “Type=simple” service, as such, dependant units will start as soon as the C* process starts rather than when C* is accepting client connections. I’d like to transition to something more complex so I can start to write additional units that depend on C*. I’ve successfully managed to set the service type to “notify” and modify C* to call sd_notify() when is ready to accept client connections. Further experimentation reveals that this is not an ideal solution. C* can take a long time (minutes to _hours_) to reach the point where it will accept client connections/queries. The default startup timeout is 90s, which causes the service to be marked failed if exceeded, hence C*, with its long startup times, will often never get the chance to transition to “active”. Part of the issue for me is trying to define what “active” means. The man pages, for “Type=forking" services, says: "The parent process is expected to exit when start-up is complete and all communication channels are set up”. I’m assuming for “notify” services, sd_notify() should be called when "start-up is complete and all communication channels are set up”. Even if this takes hours? Cassandra exposes a number of inet ports of interest: - Client connection ports for running queries via Cassandra Query Language (CQL)/Thrift (RPC) — this is what most clients use to query the database (i.e., to run `SELECT * FROM …` style queries) - JMX (Java Management Extensions) for performing management operations — the C* and 3rd-party management tools use this to call management functions and to collect statistics/metrics about the JVM and C*. The JMX socket is available a few seconds after the process is running. The CQL/Thrift ports can take far longer to become available — sometimes hours after the process starts. Cassandra only starts listening on these ports once it has joined the cluster of nodes & has synchronised its state. State synchronisation may require bootstrapping & copying large amounts of data across the network and hence take a long time to complete. Currently my dependent C* client units simply spin-wait, attempting to establish a connection to C*. This seems like duplicated effort and makes these services more complex than they need to be. My original thought was to just disable the startup timeout on the C*, but that means the unit will stay “activating” for a long time. Also means that JMX clients, which can establish connections almost immediately, would have their startup deferred unnecessarily. Ideally I’d like to be able to write units that can depend on individual ports being available from a process — i.e, when the CQL port is available, start the client unit(s) and when JMX is available, start a monitoring service. Is this possible with systemd? Alternatively, I was thinking that I could write some kind of simple process/script that attempts a connection, and exits with failure if the connection cannot be established, or success if it can. I’d then write a unit file, e.g. `cassandra-cql-port.service`: [Unit] # not really sure what combo of Wants/Requires/Requisite/BindsTo/PartOf/Before/After is needed Requisite=cassandra.service [Service] Type=oneshot RemainAfterExit=true ExecStart=/opt/bin/watch-port 9042 Restart=on-failure RestartSec=1min StartLimitInterval=0 My client units could then want/require this unit. Is this a valid approach? Or am I walking down the wrong path to use systemd to manage this? Regards, Adam ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/systemd-devel
Re: [systemd-devel] Restarting service starts stopped dependants
> On 10 Feb 2015, at 6:46 am, Andrei Borzenkov wrote: > > B is started because C is started so it is not relevant. Yes, this is expected because B is a dependency of C. Starting C should start B. I mentioned it because, to me, it was unexpected that /re/starting A caused B to start. > C is started probably due to logic "if A and C are running and you > restart A you should also restart C". I am not convinced it is the > right thing to do TBH, but well … Restarting C as a result of restarting A is correct behaviour to me, but only if C was already running. A restart = stop + start in a single transaction, correct? As such, `systemctl restart A` should: - Stop A - Stop C (if it’s running) as C can’t run without A (Requires=A) - Start A - Start C, but only if C was previously running. Starting C only if it was running previously makes sense, because `start` is not transitive. > Does patch below help? > > diff --git a/src/core/transaction.c b/src/core/transaction.c > index b0b3d6b..8901119 100644 > --- a/src/core/transaction.c > +++ b/src/core/transaction.c > @@ -1027,7 +1027,7 @@ int transaction_add_job_and_dependencies( > if (type == JOB_STOP || type == JOB_RESTART) { > > SET_FOREACH(dep, > ret->unit->dependencies[UNIT_REQUIRED_BY], i) { > -r = transaction_add_job_and_dependencies(tr, > type, dep, ret, true, override, false, false, ignore_order, e); > +r = transaction_add_job_and_dependencies(tr, > type == JOB_RESTART ? JOB_TRY_RESTART : type, dep, ret, true, override, > false, false, ignore_order, e); > if (r < 0) { > if (r != -EBADR) > goto fail; I will try. I note that TRY_RESTART “… does nothing if units are not running.” (systemctl(1)), which is perhaps the solution to the problem. Regards, Adam ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/systemd-devel
[systemd-devel] Restarting service starts stopped dependants
Hi list, Is the following expected behaviour? Given services A, B, and C. C "Requires" A & B. All units are inactive. `systemctl restart A` activates C, which in turn activates B. Whereas `systemctl start A` activates A and only A. C and B are left inactive. I can understand that starting C should start both A and B as they are Required dependencies. But I can’t grasp why *restarting* A should start C, if C is stopped. And then by extension why B should start. See below for example commands & output. Regards, Adam core@core-01 /etc/systemd/system $ find . -maxdepth 1 \( -name A.service -or -name B.service -or -name C.service \) -print -exec cat {} \; ./A.service [Unit] [Service] ExecStart=/bin/bash -c 'while true; do echo this is service A; sleep 1; done' [Install] WantedBy=multi-user.target ./B.service [Unit] [Service] ExecStart=/bin/bash -c 'while true; do echo this is service B; sleep 1; done' [Install] WantedBy=multi-user.target ./C.service [Unit] Requires=A.service B.service [Service] ExecStart=/bin/bash -c 'while true; do echo this is service C; sleep 1; done' [Install] WantedBy=multi-user.target core@core-01 /etc/systemd/system $ systemctl status --lines=0 A B C ● A.service Loaded: loaded (/etc/systemd/system/A.service; disabled; vendor preset: disabled) Active: inactive (dead) since Mon 2015-02-09 07:12:45 UTC; 8min ago Main PID: 20871 (code=killed, signal=TERM) CGroup: /system.slice/A.service ● B.service Loaded: loaded (/etc/systemd/system/B.service; enabled; vendor preset: disabled) Active: inactive (dead) since Mon 2015-02-09 07:12:45 UTC; 8min ago Main PID: 20867 (code=killed, signal=TERM) CGroup: /system.slice/B.service ● C.service Loaded: loaded (/etc/systemd/system/C.service; enabled; vendor preset: disabled) Active: inactive (dead) since Mon 2015-02-09 07:12:45 UTC; 8min ago Main PID: 20866 (code=killed, signal=TERM) CGroup: /system.slice/C.service core@core-01 /etc/systemd/system $ sudo systemctl restart A core@core-01 /etc/systemd/system $ systemctl status --lines=0 A B C ● A.service Loaded: loaded (/etc/systemd/system/A.service; disabled; vendor preset: disabled) Active: active (running) since Mon 2015-02-09 07:21:31 UTC; 1s ago Main PID: 21196 (bash) CGroup: /system.slice/A.service ├─21196 /bin/bash -c while true; do echo this is service A; sleep 1; done └─21201 sleep 1 ● B.service Loaded: loaded (/etc/systemd/system/B.service; enabled; vendor preset: disabled) Active: active (running) since Mon 2015-02-09 07:21:31 UTC; 1s ago Main PID: 21194 (bash) CGroup: /system.slice/B.service ├─21194 /bin/bash -c while true; do echo this is service B; sleep 1; done └─21200 sleep 1 ● C.service Loaded: loaded (/etc/systemd/system/C.service; enabled; vendor preset: disabled) Active: active (running) since Mon 2015-02-09 07:21:31 UTC; 1s ago Main PID: 21195 (bash) CGroup: /system.slice/C.service ├─21195 /bin/bash -c while true; do echo this is service C; sleep 1; done └─21202 sleep 1 core@core-01 /etc/systemd/system $ systemctl --version systemd 218 -PAM -AUDIT -SELINUX +IMA -APPARMOR +SMACK -SYSVINIT +UTMP +LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL -XZ -LZ4 +SECCOMP +BLKID -ELFUTILS +KMOD -IDN core@core-01 /etc/systemd/system $ sudo systemctl stop A B C core@core-01 /etc/systemd/system $ sudo systemctl start A core@core-01 /etc/systemd/system $ systemctl status --lines=0 A B C ● A.service Loaded: loaded (/etc/systemd/system/A.service; disabled; vendor preset: disabled) Active: active (running) since Mon 2015-02-09 07:35:17 UTC; 1s ago Main PID: 23972 (bash) CGroup: /system.slice/A.service ├─23972 /bin/bash -c while true; do echo this is service A; sleep 1; done └─23974 sleep 1 ● B.service Loaded: loaded (/etc/systemd/system/B.service; enabled; vendor preset: disabled) Active: inactive (dead) since Mon 2015-02-09 07:35:13 UTC; 4s ago Process: 21194 ExecStart=/bin/bash -c while true; do echo this is service B; sleep 1; done (code=killed, signal=TERM) Main PID: 21194 (code=killed, signal=TERM) CGroup: /system.slice/B.service ● C.service Loaded: loaded (/etc/systemd/system/C.service; enabled; vendor preset: disabled) Active: inactive (dead) since Mon 2015-02-09 07:35:13 UTC; 4s ago Process: 23830 ExecStart=/bin/bash -c while true; do echo this is service C; sleep 1; done (code=killed, signal=TERM) Main PID: 23830 (code=killed, signal=TERM) CGroup: /system.slice/C.service ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/systemd-devel