text
stringlengths
0
1.8k
[1839.44 --> 1842.02] prometheus or maybe an
[1842.02 --> 1843.10] open observability
[1843.10 --> 1844.50] backend that support
[1844.50 --> 1846.30] both traces on a
[1846.30 --> 1847.86] matrix elastic being one
[1847.86 --> 1849.28] i work for them but you
[1849.28 --> 1850.28] will find many other
[1850.28 --> 1852.04] vendors who also can
[1852.04 --> 1853.50] consume all observability
[1853.50 --> 1854.68] signals what about
[1854.68 --> 1856.60] auto cli from equinix
[1856.60 --> 1858.28] labs how could we use
[1858.28 --> 1859.54] that to get some open
[1859.54 --> 1860.90] telemetry in ci cd
[1860.90 --> 1861.84] systems that maybe don't
[1861.84 --> 1863.08] support it it's possible
[1863.08 --> 1864.90] that's a great point as
[1864.90 --> 1865.62] there were two
[1865.62 --> 1866.86] initiatives that comes to
[1866.86 --> 1867.80] my mind i think the
[1867.80 --> 1868.94] first one i saw what
[1868.94 --> 1869.98] came from honeycomb
[1869.98 --> 1871.40] where they created a
[1871.40 --> 1873.02] small cli to instrument
[1873.02 --> 1874.60] some ci platform
[1874.60 --> 1875.72] where the platform
[1875.72 --> 1876.50] itself didn't
[1876.50 --> 1877.52] instrument with hotel
[1877.52 --> 1879.30] otherwise if you are on
[1879.30 --> 1880.28] github actions for
[1880.28 --> 1881.64] example or maybe
[1881.64 --> 1883.44] gitlab ci you would
[1883.44 --> 1885.72] use hotel cli as
[1885.72 --> 1886.84] maybe a wrapper when
[1886.84 --> 1887.84] you invoke your maven
[1887.84 --> 1889.56] build as a wrapper when
[1889.56 --> 1890.82] you use you invoke your
[1890.82 --> 1892.62] meg file also something
[1892.62 --> 1894.34] even when you are inside
[1894.34 --> 1896.22] jenkins inside the ci
[1896.22 --> 1897.06] platform that is
[1897.06 --> 1898.26] instrumented with hotel
[1898.26 --> 1899.90] traces it's still very
[1899.90 --> 1901.98] interesting to get more
[1901.98 --> 1903.30] granularity in let's say
[1903.30 --> 1904.92] meg file because you
[1904.92 --> 1906.38] discuss a lot of meg files
[1906.38 --> 1908.06] in shipit if you want
[1908.06 --> 1909.20] granularity on what's
[1909.20 --> 1910.02] happening in your meg
[1910.02 --> 1911.14] file you can in your
[1911.14 --> 1913.10] meg file drop some calls
[1913.10 --> 1915.96] using the open cli tool
[1915.96 --> 1917.00] so that you get final
[1917.00 --> 1917.98] granularity in your
[1917.98 --> 1919.16] pipeline execution i'm
[1919.16 --> 1920.10] probably a bit lazy
[1920.10 --> 1921.40] because i just replace
[1921.40 --> 1922.86] a shell on my agents
[1922.86 --> 1924.50] so i modify shell on the
[1924.50 --> 1925.50] docker images and
[1925.50 --> 1927.08] hotel cli is enabled by
[1927.08 --> 1928.04] default there for
[1928.04 --> 1928.80] screens okay
[1928.80 --> 1929.32] interesting
[1929.32 --> 1930.68] hackish but it works
[1930.68 --> 1931.64] do you have an example
[1931.64 --> 1932.20] of how you do that
[1932.20 --> 1932.82] that's very interesting
[1932.82 --> 1933.56] i would like to check it
[1933.56 --> 1934.68] out the code i don't
[1934.68 --> 1936.02] have a code with me but
[1936.02 --> 1937.34] yeah basically you can
[1937.34 --> 1938.10] just take open
[1938.10 --> 1939.66] telemetry you create a
[1939.66 --> 1941.02] shell wrapper which
[1941.02 --> 1942.56] just sends all the
[1942.56 --> 1943.60] command worked in this
[1943.60 --> 1944.28] shell to open
[1944.28 --> 1945.80] telemetry and that's
[1945.80 --> 1946.62] it okay it's wrapper
[1946.62 --> 1947.66] which irritates all the
[1947.66 --> 1948.68] environment and which
[1948.68 --> 1950.12] is pretty transparent to
[1950.12 --> 1951.50] your system as long as
[1951.50 --> 1952.56] you use field scripts
[1952.56 --> 1954.22] obviously if you use a
[1954.22 --> 1955.70] mix let's say of bash
[1955.70 --> 1957.12] python etc then you
[1957.12 --> 1957.92] will have to instrument
[1957.92 --> 1959.66] all of these tools which
[1959.66 --> 1960.88] becomes a bit tricky but
[1960.88 --> 1961.46] still possible
[1961.46 --> 1963.20] you say cyril in one of
[1963.20 --> 1964.66] your talks that jenkins
[1964.66 --> 1966.08] in production is hard and
[1966.08 --> 1967.06] i know a thing or two
[1967.06 --> 1968.16] about that because many
[1968.16 --> 1969.02] years ago we used to
[1969.02 --> 1970.70] pair on getting cloud